Premise:
Let G be the mathematically generalized rank of the collection of all feasible and veridical aspects of multi-dimensional entities being taken into consideration. Then, all ecological embeddings (symbol-meaning bindings) that describe those entities and relationships among them, have geometric properties. Consequently, algebraic operations on any measurable quantities are feasible within structured subdomains of maximum rank G. Let E denote the space of all such ecological embeddings; G = rank(E).
Flux denotes the rate of information transfer across a surface within E, in an information theoretic sense.
Dimensionality, Size, and Degrees of Freedom:
A system S is defined as a triplet (N, R, G) such that:
N and R may each be empty. The empty system (N = Ø, R = Ø) is valid — it is inert or idempotent or void in its informational content, but not ill-formed.
A node n in N may hold a reflexive relationship (n, n) in R. In this case, n is simultaneously the sender and receiver of its own signal, ecologically coupled to itself via E. This is the minimal non-degenerate system: a single node with memory of itself.
Information transfer within S is possible if and only if R ≠ Ø and |N| ≥ 1. A system with nodes but no relationships is degenerate with no information transfer channel.
Memory of S exists if and only if E is non-trivially structured (G > 0) — nodes in N have spatio-temporal adjacency within E, and E mediates at least one relationship in R. A system with no ecological embedding, or with an unstructured one, has no memory even if N and R are non-empty.
The cost of forgetting within S depends on the ecology encoded in E — the individual, organizational, cultural, and environmental context in which S is embedded. Where relationships in R are non-linear and observer-constituted, forgetting may be irreversible. Where they are linear and observer-independent, forgetting is recoverable from residual components or external records.
If R ≠ Ø, then G > 0. Non-empty relationships require a structured embedding space to mediate them; an unstructured E cannot make relationships in R persistent or meaningful. A system with relationships but no structured ecological embedding is self-contradictory under this definition.
Note — *Constraints 4 and 6 are complementary, not redundant. Constraint 4 addresses the conditions under which memory exists within S: E must be non-trivially structured and mediating at least one relationship. Constraint 6 addresses the structural precondition for R itself: non-empty relationships cannot exist without a structured E to mediate them. Constraint 6 therefore underlies Constraint 4 — it scopes the precondition for S; Constraint 4 specifies what holds when the precondition is met.*
| N | R | E | Name | Status |
|---|---|---|---|---|
| Ø | Ø | — | Empty or void system | Valid. Informationally inert and idempotent. |
| ≠ Ø | Ø | — | Degenerate system | Valid. No transfer possible. No memory. |
| {n} | {(n,n)} | Structured | Minimal system | Valid. Single node, reflexive relation, self-memory via E. |
| ≠ Ø | ≠ Ø | Unstructured | Transfer-capable, memoryless | Formally excluded, R ≠ Ø requires G > 0 (Constraint 6). |
| ≠ Ø | ≠ Ø | Structured | Fully realized system | Valid. Transfer and memory both available. |
Any system S can be represented as a graph where nodes in N are vertices and relationships in R are edges, including self-loops. An equivalent representation is a dictionary where keys are nodes in N and values are the sets of nodes related to that key via R. G is the rank of the ecological embedding in which that graph is physically instantiated and temporally persistent. Without G > 0, such a graph would be an abstract object with no “memory”.
The formal constraints of S = (N, R, G) have a direct physical interpretation for any agent — biological, mechanical, or synthetic — capable of acting in the world.
An embodied agent is a node n ∈ N embedded in a structured E (G > 0). Its operational capacity depends on sustaining the relationships in R that allow it to transfer information via transport of energy or substance among other connected nodes in its environment. The formal constraints map to survival conditions as follows:
Recharging as a structural act. When an embodied agent’s operational capacity approaches the minimum flux threshold of its current subdomain — i.e. the subdomain can no longer support the energy transduction rate required to sustain R — the agent must migrate to a subdomain with greater flux capacity or higher Degrees of Freedom (DoF). This migration is itself a relationship in R, mediated by E. Recharging is not a special case outside the system definition; it is an instance of Code 2: a node enacting a relationship with a new node (a power source, a food supply, a charging station) within an E that makes that relationship feasible at the required flux rate.
The survival imperative follows directly: a proper agent must dynamically and creatively find and sustain the relationships — across whatever subdomains of E are accessible — that keep R ≠ Ø and G > 0. See Concept of System of Systems for the situated system framework in which this imperative is fully expressed.
definition-system.md v1.3.7 — FINAL — Human Approved