Contribution
A programmatic paper in a special issue that proposes principles for a theory of organisms. It argues against “genocentrism”, the view that genetic information determines biological organization (§1), and for organicism. It surveys the history of the idea of organization from Kant to Kauffman (§3), restates closure of constraints as the paper LIT-582 gave it (§4), and draws two implications for the relation between stability and variation (§5). Its novelty relative to LIT-582 is the status it gives closure: not a model or a criterion for drawing boundaries, but a theoretical principle.
Key insight
“All biological organisms … meet two general principles without exceptions: they are organized, and their organization undergoes variation” (p. 2). The principles are “overarching hypotheses that frame the intelligibility of the objects within the biological domain”, analogous to space and time, or to “positing a priori the phase space”, in physics (p. 2). So the claim is not that closure has been found in organisms by observation. It is that biology should take it as constitutive of its object.
Assumptions
- Thermodynamic openness. “Only thermodynamically open systems can possibly comply with the organization principle, although not any thermodynamically open system does” (p. 6).
- The process/constraint distinction (§4.1). Given a process A ⇒ B, C is a constraint on it at time scale τ iff (I) the process with C differs from the process without it at τ, and (II) the aspects of C that matter are conserved through the process at τ. “A given entity cannot be qualified as a constraint in abstracto” (p. 7). Conditions I and II must hold at the same τ.
- Direct dependence (footnote 15). C₁ depends on C₂ when C₂ constrains, at τ₂, a process that produces C₁, and no other process constrained by C₂ intervenes in time. Otherwise “the concept of dependence would include many biologically irrelevant situations”.
- Generative, not modulatory (footnote 17). Only dependences that produce the relevant aspects of a constraint count towards closure.
Key results
These are definitions and arguments, not theorems.
- Closure (§4.2, p. 8). A set C of constraints realizes closure if every Cᵢ in it depends directly on at least one other member and has at least one other member depending on it. “The principle of organization states that biological systems realize a closure of constraints” (p. 7). Further specifications are deferred to Montévil and Mossio 2015 (footnote 16). The abstract diagram: of C₁ to C₅, C₁–C₄ are dependent and C₂–C₅ generative, so {C₂, C₃, C₄} realizes closure. C₅ is an external constraint the closure is “always realized with respect to”.
- Functions (p. 9). “Constraints subject to closure define biological functions”; “constraints do not exert functions when taken in isolation”.
- Stability (§5.1, pp. 9–11). Each constraint is conserved only at its own scale τ and must be regenerated beyond it. Because of closure, its regeneration depends on other constraints, so closure maintains the constraints beyond their time scales. This is “a kind of ‘organizational inertia’”. Conservative stabilization removes a local variation only “in those cases in which local variation does not affect the constraints in charge of generating the one … being affected” (p. 11). DNA’s stability is “a consequence of the biological activity of the whole organization” (p. 11), not the source of it.
- Variation (§5.2, p. 12). When a variation does feed back into its generators, it propagates, and “two outcomes are possible. Either the resulting system cannot realize closure anymore, in which case it is not viable and disintegrates. Or, it does realize a new closure through cumulative stability”. By contrast, in a rock, a flame or a table “a local variation cannot induce other local variations, which would result in a global reconfiguration … no global variation results from local variation” (p. 12). This is offered as a reading of Darwin’s “correlated variations”.
- Scope. Biology “is the science of systems meeting the principles of organization and variation, organisms being a specific, particularly relevant, class of biological systems”; ecosystems might qualify (p. 2).
Claims
| id | claim | strength | support |
|---|---|---|---|
| C1 | Organization as closure of constraints should be adopted as a theoretical principle of biology, a priori, alongside variation | weak (programmatic) | argument by analogy with physics’ principles; no derivation from it shown here |
| C2 | Constraints subject to closure define biological functions | moderate | cited to Mossio et al. 2009; restated |
| C3 | Closure grounds biological stability by regenerating constraints beyond their time scales | weak | informal argument, examples (Na⁺/K⁺ pump, vasculature, gut); no model |
| C4 | Conservative stabilization requires that the variation not reach the constraint’s generators; otherwise variation propagates to disintegration or a new closure | moderate as a conditional | follows from the definitions if regeneration reproduces whatever its generators specify; not modelled |
| C5 | Non-organized systems (rock, flame, table) cannot turn a local variation into a global reorganization | weak | three examples, asserted |
| C6 | Genes do not determine organization; their expression presupposes it | weak here | argued by reference to other work |
Concepts
- Constraint — an entity that acts on a process at time scale τ while its relevant aspects are conserved at τ (§4.1).
- Dependence — C₁ is produced by a process constrained by C₂; C₂ is generative for C₁ (§4.2).
- Closure — mutual dependence in the sense above; the constraints in it are constitutive, the rest external (footnotes 17, 21).
- Theoretical principle — an overarching hypothesis framing the intelligibility of a domain’s objects (§1).
- Conservative / cumulative stability — removing variation versus integrating it into a new closure (§5).
Connections
- LIT-582 (Montévil and Mossio 2015, read) is the formal source; this paper adds no formal content to its definitions and drops its “tendency to closure” measure. TERM-070 holds the definition.
- The companion on variation, Montévil et al. 2016 in the same issue, is not held. Soto et al. 2016’s default state (proliferation with variation and motility) is discussed in the conclusion as connected to organization, not proposed here as a third principle.
- Rosen’s closure to efficient causation, Kauffman’s work–constraint cycles, Piaget, autopoiesis (THEORY-666), and Ruiz-Mirazo and Moreno’s basic autonomy are surveyed (§3). Footnote 9 uses the hydrological cycle as a mutual dependence that is not biological organization, because the paper requires constraints, not just processes.
Bearing on the record
- The registration’s summary needs a correction. It says the paper proposes organization “alongside variation and the default state of proliferation with variation”. The paper proposes two principles, organization and variation; the default state is Soto et al.’s, and the conclusion only relates organization to it.
- TERM-070. Consistent; the same definition with the same direct and generative qualifications (footnotes 15, 17). The paper adds the explicit statement that a closure is always realized relative to external constraints (C₅), which TERM-070 could note.
- CLAIM-516. The paper supports two parts of the claim. A dependence cycle is needed for anything to persist beyond its own time scale. And a cycle of processes alone, like the hydrological cycle, is not closure. It says nothing quantitative.
- CLAIM-521 and CLAIM-522. The paper’s stability argument is exactly the inference these claims caution against: it reads stability off the network of dependences without a dynamical model. Its own condition for conservative stabilization (that the variation not reach the generators) is a structural condition. Whether the regeneration recovers is a separate dynamical question that CASE-189 shows can fail at a saddle-node while the coupling looks healthy.
- THEORY-514 (Baedke): the “new closure” outcome of §5.2 is the positive mark Baedke proposes for organisms. Mossio et al. state it as a feature of all organized systems, with no claim that only organisms achieve it. Their scope (biology covers any system meeting the principles, ecosystems possibly included) agrees with Baedke’s negative half.
- THEORY-072. Supports its demarcation from the flame (pp. 6, 12) but adds no argument beyond LIT-582.
- No instruction for ML practice.
Limitations
- No model, measurement or proof; the stability and variation claims are argued from examples.
- The principle’s a priori status is asserted by analogy with physics; what would count against it is not said.
- “Without exceptions” (p. 2) is a definitional claim once biology is defined as the science of systems meeting the principles, so it cannot be tested by finding an unorganized organism.
- The contrast with rocks, flames and tables is three examples.
Open questions
- What dynamical condition on a closure network makes conservative stabilization succeed? The structural condition is necessary, not sufficient.
- Can the claim that only closed systems turn local into global reorganization be stated so that it could fail, for example for coupled non-closed dissipative structures?