Part 1 established the skeleton: an entity is anything with a describable state - anything you can assign an $X$ to - and auto-causal density $\rho_{\text{ac}}$ measures how strongly it sustains itself. A strong entity is a structure with high auto-causal density $\rho_{\text{ac}}$ . Its generalized mass $\mathcal{M} = \int \rho_{\text{causal}} \, d\mu$ measures total causal content, its momentum is $p = \mathcal{M}\dot{X}$ , and it is subject to forces $F = \mathcal{M}\ddot{X} + \dot{\mathcal{M}}\dot{X}$ . Entities differ in how strongly they couple to various domains - captured by a scalar $\kappa$ or a coupling tensor $T^i{}_j$ . Fields $\Phi(x)$ assign force values across domains; entities at different positions experience different forces even with identical coupling structure.

Epimechanics handles most entities straightforwardly. A rock, a person, a protein, a market price - each has a state, a state space, and coupling to fields (with varying degrees of auto-causal density). But some structures resist this description. They are distributed. They have no single location. They persist after their original instantiation dissolves. They exert causal force through mechanisms no individual part possesses. The scientific method is not a person or a text. A religion is not its founder or its scripture. A market is not any of the traders who constitute it. What are these things?

The question is: when does an aggregate of entities develop its own auto-causal density - its own self-sustaining loop - rather than depending entirely on the loops of its parts?

1. When Does an Aggregate Become an Entity?

Three rigorous answers exist in the literature, from three different theoretical traditions. They are compatible and complementary.

1.1 Causal Emergence (Hoel, Albantakis & Tononi)

Hoel, Albantakis & Tononi (PNAS, 2013) provide the most formally precise answer. Their criterion is effective information (EI): the amount by which a mechanism constrains its effects. A description earns causal standing when $EI(\text{macro}) > EI(\text{micro})$ .

The mechanics: consider a system with micro-level components (neurons, people, particles) and a potential macro-level description (brain region, society, gas). At the micro level, each mechanism has a certain determinism (how precisely it constrains its outputs) and a certain degeneracy (how many different inputs produce the same output). Micro-level systems are often highly degenerate - many configurations of neurons produce the same neural output, many configurations of individuals produce the same collective behavior. When coarse-graining to the macro level reduces degeneracy while preserving determinism, the macro description captures more causal structure per unit of description than the micro does.

The result: "Although the macro level supervenes upon the micro, it can supersede it causally, leading to genuine causal emergence."

Applied here: a religion earns its own state variable X when describing it at the macro level - membership dynamics, institutional behavior, theological attractors, conflict with competing systems - captures more causal structure than describing its individual adherents. The adherents are the substrate. The religion is causally real at a higher level of description. This is an information-theoretic criterion.

The same logic applies to markets, nations, languages, ideologies, and scientific fields. Each earns entity status - or fails to - by whether macro-level description outperforms micro-level description on EI.

1.2 Near-Decomposability (Simon)

Herbert Simon ("The Architecture of Complexity," Proceedings of the American Philosophical Society, 1962) gives the dynamical-timescale criterion: a system forms genuine sub-units when within-unit coupling $\gg$ between-unit coupling.

The key claim: when the short-run behavior of each subsystem is approximately independent of other subsystems, the subsystem becomes functionally real. The tight internal coupling creates a causal envelope around the unit; the loose external coupling means external disturbances take long to penetrate. On the timescale of within-unit dynamics, the unit behaves as a semi-autonomous entity.

Simon's examples span biology (cells within organisms, organs within bodies), organizations (departments within firms), and social structures (families within communities). The criterion is quantitative: the ratio of within-unit coupling strength to between-unit coupling strength determines how cleanly the unit boundary separates causally.

Applied here: a nation-state earns entity status when the causal coupling among its citizens, institutions, legal systems, and infrastructure is significantly tighter than coupling to external systems. Where that coupling ratio is high, "the nation" is causally real. Where it is low - in ungoverned spaces, dissolved states, or extensively colonized societies - the nation-level description loses causal grounding.

1.3 Quantifying Collectivity (Daniels, Krakauer & Flack)

Daniels, Krakauer & Flack (Current Opinion in Neurobiology, 2016) operationalize entity status with two empirically measurable quantities:

Amplification: the sensitivity of macro-scale behavior to micro-scale information. High amplification means the collective does something that individual members' states do not straightforwardly predict - the macro is causally doing something the micro is not.
Decomposability: whether collective behavior is synergistic or merely additive. A collective earns unit status when its behavior is substantially synergistic - when the whole does something that the parts, analyzed individually, cannot account for.

The practical implication: synergy is measurable. Given the micro-states and the macro-states, information-theoretic measures can determine whether macro-behavior contains information not present in any linear combination of micro-states. When it does, the collective earns its own state variable.

Synthesis

These three criteria are complementary, not competing:

Hoel provides the causal-information criterion: $EI(\text{macro}) > EI(\text{micro})$
Simon provides the dynamical-timescale criterion: within-unit coupling $\gg$ between-unit coupling
Daniels/Krakauer/Flack provide the empirically measurable operational criterion: synergistic behavior

A genuine meta-entity typically satisfies all three. An apparent collective that fails all three is merely an aggregate - its description should remain at the micro level.

2. Three Substrates of Meta-Entity Existence

A meta-entity exists simultaneously across three substrates:

2.1 Physical Substrate

The physical substrate is the embodied layer: the people, objects, places, and infrastructure that carry the meta-entity's patterns. A religion exists in its practitioners' bodies, gestures, rituals, and physical communities. A scientific institution exists in its laboratories, servers, and human practitioners. A market exists in the physical locations and communication infrastructure of its participants.

The physical substrate is causally fundamental in one direction: information and mimetic patterns require physical instantiation to exist and act. But it is not determinative: the same physical substrate can carry different meta-entities depending on the informational and mimetic structure superimposed on it. A laboratory reorganized around different scientific norms becomes a different scientific institution even with identical people and equipment.

2.2 Informational Substrate

The informational substrate is the encoded layer: texts, laws, contracts, databases, institutional protocols, architectural plans, and formal knowledge structures that persist independent of any particular physical instantiation. A religion's informational substrate includes its scriptures, canon law, theological commentaries, organizational charts, and doctrinal formulations. A market's informational substrate includes its legal framework, accounting standards, regulatory structures, and pricing mechanisms.

The informational substrate confers persistence across physical turnover. Individual adherents die; the textual tradition continues. Individual traders leave; the exchange rules remain. The informational substrate is what allows meta-entities to survive the replacement of their physical components - the partial answer to the Ship of Theseus problem.

2.3 Mimetic Substrate

The mimetic substrate is the self-reproducing conceptual layer: patterns that propagate from mind to mind through imitation, teaching, and cultural transmission. Dawkins' The Selfish Gene (1976) introduced the meme as the unit of cultural replication, explicitly analogizing to genetic replication: cultural patterns that replicate with high fidelity, high fecundity, and longevity proliferate; those that replicate poorly diminish. Blackmore's The Meme Machine (1999) develops this further: the mimetic substrate has its own selection pressures, its own fitness landscape, and its own evolutionary dynamics independent of the interests of the physical hosts carrying it.

A religion's mimetic substrate includes the concepts, narratives, metaphors, emotional resonances, and behavioral patterns that practitioners acquire and transmit - including to people who never encounter the physical or informational substrate directly. The mimetic substrate is what allows a meta-entity to colonize new territory without institutional infrastructure: a single practitioner can seed the mimetic substrate in a new population.

2.4 The Coupling Among Substrates as the Source of Causal Power

The critical point is that meta-entity causal power derives not from any single substrate but from the coupling among all three substrates simultaneously.

Define $C(t)$ as the meta-entity's aggregate capacity to influence state X at time t - the meta-entity's effective coupling to the state space of the systems it acts upon. Formally, $C(t) = \sum_j \kappa_{Mj}(t) \cdot |\dot{X}_j(t)|$ - the sum of the meta-entity's coupling to each component entity $j$ times that entity's rate of state change. This measures the aggregate directed causal activity flowing through the meta-entity's coordination structure at time $t$ . When $C(t) > 0$ , the meta-entity is actively influencing state trajectories; when $C(t) \to 0$ , it has become causally inert.

$C(t)$ is a function of:

Embodiment density: how many physical entities carry the meta-entity's patterns, and how strongly each is coupled to the pattern set
Informational connectivity: how densely networked the informational substrate is - whether its texts, laws, and protocols form a coherent, cross-referencing, self-reinforcing structure or a fragmented archive
Mimetic reproduction rate: how reliably conceptual patterns replicate when transmitted, and how high the fidelity of that replication is

A meta-entity with high density in all three substrates and strong cross-substrate coupling has enormous $C(t)$ . A meta-entity with high density in one substrate but weakly coupled to the others has reduced $C(t)$ - the single substrate's causal power is not amplified by reinforcement from the others.

The causal power of a meta-entity comes from the degree to which practitioners, texts, and mimetic patterns reinforce, regenerate, and amplify each other - the coupling among substrates, rather than any single substrate.

2.5 Systemic Coupling and Part-Whole Efficiency

Auto-causality operates at specific scales of time and space. A sub-loop that sustains itself locally may not contribute to — or may actively drain — the containing meta-entity at larger scales. Not all auto-causal sub-loops increase the containing system's $\rho_{\text{ac}}$ . A sub-entity can sustain itself while:

Contributing nothing to the meta-entity (decoupled)
Draining the meta-entity's resources (parasitic)
Producing entropy without useful causal work (dissipative)

Systemic coupling $\kappa_{\text{sys}}$ measures the structural relationship between scales:

$\kappa_{\text{sys}} = \frac{\partial \rho_{\text{ac}}^{\text{meta}}}{\partial \rho_{\text{ac}}^{\text{sub}}}$

When $\kappa_{\text{sys}} > 0$ , the sub-entity's persistence contributes to the meta-entity's persistence. When $\kappa_{\text{sys}} < 0$ , it's parasitic — the sub-entity persists at the meta-entity's expense. When $\kappa_{\text{sys}} = 0$ , it's decoupled — self-sustaining but structurally disconnected from the larger scale.

Note on terminology: "Contribution" here is structural, not normative. It means: does increasing the sub-loop's $\rho_{\text{ac}}$ increase the containing system's $\rho_{\text{ac}}$ ? This describes causal relationships between scales, not value judgments. Existence itself is the selection criterion — structures that persist, persist; those that don't, don't. As noted in The Persistence Reversal: "Existence is the filter that selects for persistence mechanisms — not as a value statement but as a structural fact."

Examples across substrate types:

Substrate	Positive $\kappa_{\text{sys}}$	Negative $\kappa_{\text{sys}}$	Zero $\kappa_{\text{sys}}$
Physical	Organ serving organism	Tumor	Vestigial structure with no metabolic cost
Informational	Standard that improves coordination	Legacy system blocking migration	Archived protocol no one references
Mimetic	Practice that strengthens culture	Rumor that divides practitioners	In-group joke with no external effect

Causal efficiency $\eta$ is the fraction of sub-entity causal power that contributes to the meta-entity's persistence:

$\eta = \frac{\mathcal{P}_{\text{systemic}}}{\mathcal{P}_{\text{total}}} = \frac{\text{causal power contributing to meta-entity persistence}}{\text{total causal power of sub-entity}}$

An idling engine ( $\eta \approx 0$ ) sustains itself (combustion maintains combustion) but does no work on the containing system. A well-integrated department ( $\eta$ high) converts most of its activity into meta-entity persistence.

The temporal dimension: parasites can become organs.

Instantaneous $\kappa_{\text{sys}}(t)$ is not the full picture. A sub-entity with $\kappa_{\text{sys}}(t) < 0$ today but $d\kappa_{\text{sys}}/dt > 0$ may be integrating — borrowing from the system now to contribute later.

Historical examples:

Mitochondria: Originally parasitic bacteria that invaded host cells; now essential for eukaryotic energy metabolism. Initial $\kappa_{\text{sys}} < 0$ became $\kappa_{\text{sys}} \gg 0$ over evolutionary time.
R&D departments: Initially pure cost centers draining organizational resources; later the source of competitive advantage.
Immune system development: Costly to maintain, initially net-negative for organism survival (resource drain, risk of autoimmunity); later essential for long-term persistence.
Venture investment logic: Fund sub-entities with negative current $\kappa_{\text{sys}}$ but high expected future integration.

The relevant quantity for evaluation is not instantaneous systemic coupling but expected future contribution:

$\mathbb{E}\left[\int_t^{\infty} \kappa_{\text{sys}}(t') \cdot \rho_{\text{ac}}^{\text{sub}}(t') \, dt'\right]$

This complicates judgments about "waste" or "parasitism." A currently negative- $\kappa_{\text{sys}}$ loop may be the seed of future systemic contribution. The question is not "is this parasitic now?" but "what is the trajectory of integration?"

Failure modes:

Premature optimization: Eliminating all negative- $\kappa_{\text{sys}}$ sub-entities kills the source of future structural innovation.
Parasitic capture: Tolerating negative- $\kappa_{\text{sys}}$ loops too long allows them to grow powerful enough to resist integration or removal.
Decoupled persistence: Zero- $\kappa_{\text{sys}}$ structures that consume no resources but occupy niches, blocking potential positive- $\kappa_{\text{sys}}$ alternatives.

The meta-entity governance problem is: maintain a portfolio of sub-entities with varying $\kappa_{\text{sys}}(t)$ , tolerating temporary parasitism where integration trajectory is positive, eliminating parasitism where it's not, and avoiding decoupled lock-in.

3. Case Studies

3.1 A Major World Religion

A major world religion is all three - persons, texts and institutions, and self-propagating ideas - simultaneously, coupled across all three substrates, with the coupling among them constituting its causal power.

The physical substrate: hundreds of millions to billions of practitioners, a global infrastructure of worship spaces, educational institutions, hospitals, and media organizations, and a continuous generational transfer of practitioners through birth, socialization, and conversion.

The informational substrate: textual canons refined over centuries, systematic theology, legal codes (canon law, sharia, halakha), institutional governance structures, and educational curricula that encode the tradition with enough precision to regenerate it after disruption.

The mimetic substrate: narratives, rituals, emotional resonances, moral intuitions, and cosmological frameworks that spread through participation and imitation - patterns so well-fitted to human cognitive architecture that transmission requires minimal formal infrastructure.

David Sloan Wilson's Darwin's Cathedral (2002) provides the clearest causal-functional analysis: Calvinist Geneva is examined as a design document for a functional social organism. The theological structure, the disciplinary institutions (the Consistory), the economic ethic, and the political arrangements are analyzed not as expressions of spiritual conviction but as mechanisms that solved collective action problems and enhanced group fitness. Whether or not that was the intention is irrelevant to the function. The causal-emergence criterion is satisfied: the macro-level description of Calvinist Geneva captures causal structure - its resilience, its expansion, its internal governance capacity - that no description of its individual members would predict.

Henrich's The Secret of Our Success (2015) argues that cultural evolution - the selective transmission of practices, beliefs, and institutions - is the primary mechanism of human adaptive success. Religions are one of the most powerful vehicles for this cultural evolution: they encode adaptive behaviors in ritual form, transmit them across generations with high fidelity, and enforce them through social accountability structures. The mimetic substrate of a religion is shaped by millennia of selection pressure toward patterns that are transmissible, memorable, and behaviorally generative.

Peter Turchin's cliodynamics (Ages of Discord, 2016) demonstrates that meta-entity dynamics - including religious movements - are quantitatively predictable at civilizational scales. His Structural-Demographic Theory tracks five measurable variables: population size relative to productive capacity, elite overproduction (ratio of elite aspirants to elite positions), state fiscal health, social cohesion, and political instability. These variables form a dynamical system with secular cycles - roughly 150-year oscillations between integrative and disintegrative phases - that have been verified across multiple civilizations and time periods. Epimechanics' state variable X for a civilization is a vector in this five-dimensional space; the forces driving $dX/dt$ are the structural pressures these five variables exert on each other.

Bruno Latour's Actor-Network Theory (Reassembling the Social, 2005) adds a further dimension: non-human actants - texts, technologies, architectures, and material objects - have genuine causal agency in social networks. A scripture is not a passive record of human intentions; it is an actant that constrains, enables, and redirects the behavior of the humans who interact with it. The informational substrate is causally active, not merely archival. Latour's framework aligns with the coupling-among-substrates picture: the scripture acts on the practitioner; the practitioner generates new scripture; the institution enforces which scriptures count; the mimetic pattern selects which features of the scripture propagate. This is a closed causal loop among all three substrates.

3.2 The Scientific Method

The scientific method has extraordinarily high $C(t)$ by any measure.

Physical substrate: tens of millions of trained practitioners distributed across all major nations, sustained by universities, research institutions, corporations, and government agencies with combined annual expenditure on the order of $2 trillion globally.

Informational substrate: a self-referencing network of journals, preprint servers, patents, and educational curricula encoding methodological standards, peer review norms, statistical practices, and domain-specific knowledge with sufficient precision to regenerate scientific practice after disruption.

Mimetic substrate: the core pattern is self-reinforcing by design. Results that work propagate; results that fail die. The selection criterion is built into the mimetic pattern itself - an extraordinarily high-fidelity replication mechanism. A practitioner trained in the scientific method does not merely transmit the practices; they transmit the criterion for selecting which practices to transmit. This is the deepest layer of the mimetic substrate's fitness.

The scientific method also satisfies all three entity criteria: Hoel's EI criterion (the macro-level description of "scientific consensus" captures causal structure in predicting technological development that no description of individual scientists provides); Simon's near-decomposability criterion (within-discipline coupling far exceeds between-discipline coupling on short timescales, though cross-discipline coupling creates the meta-entity's most significant advances); and Daniels et al.'s synergy criterion (scientific consensus is synergistic - it reflects information not present in any individual's beliefs).

3.3 A Dying Language

A dying language illustrates the opposite end of the $C(t)$ spectrum and shows concretely how Epimechanics' quantities map onto existing measurement practices - suggesting these are genuine measurement problems.

Physical substrate: few remaining native speakers, aging demographic profile, limited geographic concentration, no major institutional infrastructure for transmission.

Informational substrate: possibly thin - a partial dictionary, a grammar sketch, some recorded oral literature - or possibly rich (for well-documented languages) but not actively used. A rich archive with no living practitioners is informational infrastructure with no physical substrate coupling. The cross-substrate coupling is broken.

Mimetic substrate: low fidelity transmission, few opportunities for imitation, no community of practice to enforce the mimetic patterns that constitute fluency. The reproduction rate is below replacement.

$C(t)$ is low and falling. The language is not yet gone - it still has some physical instantiation, some causal capacity to influence the state of linguistic space - but the trajectory of $dC/dt$ is clear. The entity is dissolving not because any of its three substrates have individually vanished but because the coupling among them has fallen below the threshold required to regenerate each substrate from the others.

This is a direct measurement problem. Endangered language assessment is, in these terms, the measurement of $C(t)$ and its rate of change, with intervention aimed at strengthening cross-substrate coupling before $C(t)$ falls below recovery thresholds.

4. The Extended Mind and Distributed Cognition

Epimechanics' account of meta-entities requires clarity on one philosophical pressure point: what exactly constitutes the boundary of an entity?

Clark & Chalmers' "The Extended Mind" (Analysis, 1998) made the case that mental states extend into the environment. Their canonical example: Otto has Alzheimer's and relies on a notebook to remember where the museum is. Clark and Chalmers argue that the notebook's information functions as part of Otto's cognitive system in exactly the same sense that the biological memory of a person without Alzheimer's does - it is reliably there when needed, it is consulted before action, and it guides behavior. The mind extends to wherever the relevant causal coupling exists.

Epimechanics incorporates and generalizes this: any entity's state X may extend across its physical boundary to include the informational and mimetic substrates it is tightly coupled to. Otto's state X includes the notebook. A scientist's state X includes their lab notebook, their published papers, their research community's shared knowledge. A corporation's state X includes its legal charter, its institutional memory, its contractual relationships.

The boundary of an entity is wherever coupling falls below a critical threshold $\kappa_c$ - a dynamical boundary, not a spatial one. This has several consequences:

Entity boundaries are not fixed. They expand when coupling increases (a person becomes more embedded in an institution) and contract when coupling decreases (a person leaves a community, loses a collaborator, forgets).
Entity boundaries are substrate-specific. An entity may have a tight physical boundary but an extended informational boundary - a hermit whose ideas are widely read has a narrow physical coupling and a broad informational one.
Entity boundaries are asymmetric. A's state may include B's information without B's state including A's information. Influence is not symmetric.

Bateson's Steps to an Ecology of Mind (1972) anticipated this: "The boundary of the self is where information ceases to flow." In Bateson's terms, the self extends to the limits of the information flow that constitutes it. Epimechanics operationalizes this as the $\kappa_c$ threshold on coupling strength.

For meta-entities, this means the entity's boundary is the manifold in (physical × informational × mimetic) space where the cross-substrate coupling tensor $\Gamma_{ij}$ drops below the critical coherence threshold. A religion's boundary is not its membership rolls (physical) or its canonical texts (informational) or its core beliefs (mimetic) - it is the surface in the product space of all three where mutual reinforcement falls below the level required to propagate the pattern. Formally, the boundary is the locus where $\lambda_{\min}(\Gamma) = \kappa_c$ , with $\lambda_{\min}(\Gamma)$ the minimum eigenvalue of the coupling tensor. Note that $\kappa_c$ is domain-specific and empirically determined: it represents the minimum inter-substrate coupling strength below which the meta-entity's macro-level description loses causal information relative to the micro-level description - the point at which Hoel et al.'s $EI(\text{macro}) \leq EI(\text{micro})$ .

5. Implications

5.1 The Cost of Misidentifying the Level of Analysis

Treating meta-entities as mere aggregates of individuals is causally impoverishing. Hoel's EI criterion establishes that macro-level descriptions can capture more causal structure than micro-level ones. Doing policy analysis, legal analysis, or AI analysis at the individual level when the causally active entity is a meta-entity means modeling at the wrong level of abstraction.

The practical consequences are significant:

Policy targeting: Interventions aimed at individual adherents of an extremist movement may be ineffective if the causally active entity is the movement's institutional and mimetic infrastructure. The right level is the meta-entity's cross-substrate coupling structure.
Legal standing: Law currently grants corporate personhood - the recognition that a corporation is a legal entity with standing, rights, and responsibilities distinct from its members. Epimechanics provides a principled basis for this and extends it: any system satisfying the three entity-status criteria is a causally real unit of analysis - a level of description that captures causal structure the micro-level misses. Whether current law correctly identifies which systems meet this bar is a separate question.
AI governance: The question of when an AI system becomes a meta-entity - with its own state variable, its own coupling structure, and its own causal standing - is practically consequential. It determines the appropriate level of analysis for understanding the system's causal power, predicting its behavior, and assigning accountability for its effects.

5.2 When Does an AI System Become a Meta-Entity?

An AI system with millions of users, a documented impact on user behavior, institutional integration into educational and professional workflows, and mimetic patterns that propagate through the culture it influences is plausibly approaching meta-entity status by all three criteria.

Hoel's criterion: does the macro-level description of "the AI system's influence on collective belief and behavior" capture more causal structure than the description of individual user interactions? As usage scales and systemic effects emerge, almost certainly yes.

Simon's criterion: is the within-system coupling (between model weights, training data, deployment infrastructure, and user community) substantially greater than between-system coupling? For major deployed systems, yes - the coherent behavior of a specific model's outputs across diverse interactions is a tight internal coupling.

Daniels et al.'s criterion: is the collective influence synergistic - does the AI system produce epistemic and behavioral changes that no description of individual interactions predicts? Yes, if the system's influence includes emergent cultural patterns, new linguistic norms, or shifts in collective epistemic standards.

If these criteria are met, the AI system has earned entity status. This does not settle questions of consciousness or moral standing - those are separate questions addressed in Part 3. It does mean the appropriate unit of causal analysis is the system as a whole, not the individual interaction.

When does a platform become a meta-entity? By the same criteria: when its macro-level description captures more causal structure than the description of individual users; when within-platform coupling substantially exceeds off-platform coupling; when collective behavior is synergistic. By these measures, major social media platforms are already meta-entities - their effects on political discourse, epistemic standards, and social norms are not decomposable into the sum of individual user interactions.

5.3 On Governance and Accountability

If meta-entities are genuine causal entities, then accountability frameworks must address them at the meta-entity level. Holding only individuals accountable when the causally active entity is an institution can miss the structural drivers of behavior - just as describing an electromagnet solely in terms of individual electron trajectories misses the macro-level field dynamics that determine the system's effects. The electrons are the substrate; the electromagnet is the causally relevant entity.

This has direct implications for:

Corporate accountability: Legal structures that distribute liability across individual officers may fail to address institutional behavior that no individual directed.
Religious institutions: Cases where institutional patterns - of abuse, of financial misconduct, of political interference - cannot be attributed to any individual actor require accountability frameworks operating at the meta-entity level.
AI systems: If an AI system is a meta-entity with its own causal standing, accountability for its effects cannot be attributed solely to its designers, deployers, or users. The system, as an entity, bears some form of causal responsibility - which requires frameworks for addressing that responsibility that do not yet exist.

Closing

Epimechanics now has two layers: the generalized mechanics of Part 1 ( $X$ , $\dot{X}$ , $\mathcal{M}$ , $p = \mathcal{M}\dot{X}$ , $F$ , $W$ , $T^i{}_j$ , $\Phi$ ) and the meta-entity theory of this part (three substrates, three entity criteria, distributed coupling as the source of causal power).

Together they describe how higher-level auto-causal loops emerge from lower-level ones - how the mutual coupling of practitioners, informational infrastructure, and mimetic patterns can produce a new region of elevated $\rho_{\text{ac}}$ at the macro scale, with its own state variables, its own mass, its own local time. The meta-entity is the emergent auto-causal loop - and by the three criteria above, that emergence is causal.

The next question follows naturally: what do these entities - individual and meta - know? How do they model the world? What are they trying to do? That requires the concepts of intelligence, consciousness, and agency.

← Part 1: Generalized Mechanics | → Part 2.5: Entity Interaction | → Part 3: Intelligence, Consciousness, and Agency

Epimechanics - Part 2: Meta-Entities

Contribute to Epiphysics