The campfire
The simplest closure: a campfire.
Fire produces heat. Heat breaks down fuel. Fuel breakdown produces volatile gas. Volatile gas ignites. Ignition produces fire. The fire makes the conditions that sustain the fire. Every product in the loop is produced by something else in the loop.
That's catalytic closureA set of chemical reactions where every catalyst needed is produced by the reactions themselves. No external catalyst required. The mathematical definition of self-sustenance.. Not a metaphor. A structural property: a set of processes where every product is sustained by the set itself.
What closure means precisely
A Reflexively Autocatalytic and Food-generated setA network satisfying two conditions: (1) every reaction is catalyzed by something the network produces, and (2) every reactant can be built from simple ‘food’ molecules via network reactions. (RAF) has two requirements:
Reflexively autocatalytic
Every reaction in the set is catalyzed by at least one molecule produced by the set. Nothing depends on an outside catalyst.
Food-generated
Every reactant can be built from a small "food set" of raw materials through reactions in the set. You can bootstrap the whole thing from simple inputs.
The campfire, a metabolic network and the cardiac calcium signaling cycle all satisfy the RAF conditions. The mathematical structure is identical.
The key insight: closure is a network propertyA property of the whole system, not any individual component. You can’t point to one reaction and say ‘that’s the closure.’ Remove any single reaction and the closure may still hold., not any individual reaction. You can't point to one reaction and say "that's the important one." Remove any single reaction and the RAF might still hold. Closure is a global property of the set.
Your body is an RAF
Your cells are full of enzymesProteins that catalyze (speed up) chemical reactions without being consumed. Your body contains thousands of different enzymes, each catalyzing specific reactions.. Enzymes catalyze reactions. Those reactions produce (among other things) more enzymes. Enzymes → reactions → enzymes. Catalytic closure.
Hordijk and Steel showed this computationally1: real metabolic networks satisfy the RAF property. Not approximately. Exactly. Every reaction catalyzed, every reactant derivable from food molecules.
This is why you don't need to be plugged in. A car engine needs someone to supply fuel and maintain parts. Your body maintains itself from food + air + water. That self-maintenance is catalytic closure.
The cardiac RAF
Our analysis finds that the heart's calcium signaling cycle satisfies the RAF property structurally, not just as a metaphor.◆ our resultAn original, not-yet-peer-reviewed result from this research program — distinct from established consensus. The specific figures below (17/17 reactions, zero essential reactions) are this program's own computation on a model of the cardiac calcium cycle.
A caveat that applies to every claim of this shape. Binary RAF existence — the yes/no question “is this network an RAF?” — is generic at moderate catalysis density. Randomly generated networks acquire the property routinely once enough reactions are catalyzed, so answering “yes” is close to no evidence at all. The informative quantity is not existence but the fraction of the reaction set that remains non-zero under perturbation — which is why the number worth reporting below is 17/17 with zero essential reactions, and not the bare fact that an RAF exists.
In our model, 17 out of 17 reactions in the cardiac calcium cycle are part of the RAF, with zero essential reactionsReactions whose removal would break the RAF property. Zero essential reactions means maximum robustness — the system tolerates any single failure. — meaning you can remove any single reaction and the RAF property still holds. On this analysis the system is maximally robust.
This is what makes the heart different from a machine. A car engine has essential parts — remove the spark plug and it stops. The cardiac RAF has no single essential reaction. It sustains itself through redundant closure.
Two depths of closure
Depth 1: Self-sustenance
The system makes everything it needs to keep running. A campfire sustains itself. A metabolic network sustains itself. The RAF property: every catalyst produced internally.
Systems with self-sustenance: campfire, metabolism, ecosystem nutrient cycling, cardiac calcium signaling.
Depth 2: Self-creation (autopoiesisSelf-creation. A system that produces its own boundary and internal components. Coined by Maturana and Varela (1971). The distinction between a machine (made by something else) and a living system (makes itself).)
The system doesn't just maintain its parts — it creates its own boundary. It produces the container that separates inside from outside. The cell membrane is produced by the metabolism it contains.
This is the jump from "makes its parts" to "makes itself as a distinct entity." Few systems reach depth 2. The cardiac system does: the heart generates the electrical and mechanical conditions that maintain its own tissue boundary.
Most biological systems achieve depth 1 (self-sustenance). Depth 2 (self-creation) is rarer and more interesting. It's the difference between a factory that makes its own parts and a factory that also builds and maintains its own walls.
Eight formalizations, six discoveries
The same mathematical structure was discovered independently at least six times across four countries and fifty years:
(M,R)-systemsRobert Rosen’s (1958) category-theory formalization of life. M = metabolism, R = repair. Showed life is algebraic closure, not mechanism.. Category theory formalization. USA. Showed life = algebraic closure, not mechanism.
Autopoiesis. Self-creating networks. Chile. Defined life as self-boundary-production.
Chemoton. Minimal self-reproducing chemical system. Hungary. Independently of all the above.
HypercyclesThe hypercycle — coupled autocatalytic cycles where each cycle catalyzes the next — was named and developed by Eigen & Schuster (1977–78). A precursor to RAF theory.. Coupled autocatalytic sets. Germany. Origin of life as mutual catalysis.
Social autopoiesis. Applied to communication systems. Germany. Extended the concept beyond chemistry entirely.
Autocatalytic sets. Random graph theory. USA. Showed closure arises spontaneously at sufficient complexity.
λ-chemistry. Functional closure in abstract chemistry. USA/Austria.
RAF theory. Rigorous graph-theoretic formalization. Netherlands/New Zealand. Made it computable.
Eight formalizations. Different notations, different motivations, different fields. All describe the same structural property: a network of processes that collectively produce all the catalysts they need.
This is a FACETSA cross-domain survey cataloguing how the mathematics of criticality was independently developed across roughly a dozen scientific fields. Preprint, arXiv:2601.22389, with R. Macomber — under review, not yet peer-reviewed.-scale convergence. Just as ~12 fields independently developed criticality mathematics, at least 6 independent lines of work arrived at closure mathematics. The convergence is invisible because each community uses different terminology.
Why closure has no tracer molecule
Topology has φ. Criticality has power-law exponents. Computation has Turing universality proofs. Closure has... nothing equivalent.
This is because closure is a network property, not a quantity. You can't measure it with a single number. You have to analyze the full reaction network and check: is every reaction catalyzed? Is everything derivable from food? These are graph-theoreticAnalyzed using the mathematics of networks (nodes and edges). RAF theory represents reactions as nodes and catalysis relationships as edges, then checks structural properties. questions, not numerical ones.
The test for closure is structural: compute the RAF and check if it spans the system. Binary result (yes/no), but the test itself requires the full network as input.
This makes closure harder to detect "in the wild" than the other skeletons. You need the full wiring diagram. But when you have it — as with the cardiac calcium cycle — the test is unambiguous.
Recap
- Skeleton
- Catalytic / generative closure — RAF theory, autopoiesis, (M,R)-systems. A network where every process is sustained by the network itself.
- Depth 1: Self-sustenance
- The system makes everything it needs to keep running. Every catalyst produced internally. No essential components.
- Depth 2: Self-creation
- The system creates its own boundary. Not just maintaining parts — producing the container. The difference between a campfire (d1) and a living cell (d2).
- Tracer
- None. Closure is a structural property, not a numerical signature. Detection requires the full reaction network.
- Key insight
- Closure is what makes living systems different from machines. Machines have essential parts. Closed systems have no essential parts — the closure itself is the essential property.
- Convergence
- 8 independent formalizations across 6 discovery events, 4 countries, 50 years. Rosen, Maturana/Varela, Gánti, Eigen, Luhmann, Kauffman, Fontana/Buss, Hordijk/Steel.