Abstract

This paper presents a foundational proposal in which the primitive constituent of the proposed physical ontology is not an object but an act: differentiation, the elementary production of distinguishability. The paper is organized in three explicit levels. Level I states the ontological claim in prose: that becoming is prior to being, and that change, energy, and time are three descriptions of one underlying activity rather than three separate primitives. Level II gives this claim a precise mathematical skeleton: admissibility rules for the primitive act, a composition law, an elementary but genuine invariance result, and a concrete, explicitly labeled candidate for the transition law that was missing from earlier drafts of this work. Level III shows how energy, time, geometry, and matter are proposed to emerge from that skeleton, including a treatment of energy in which linearity in the intrinsic measure is derived, under stated regularity assumptions, from extensivity, rather than assumed. The paper closes by separating, without blurring, what has been established by an actual argument, what is offered as an explicitly labeled proposal awaiting derivation, and what remains genuinely open. Every mathematical result in this paper is elementary. None of them is a substitute for the physical demonstration the theory still owes, which is that its proposed dynamics, run forward, produces a sparse, low dimensional, Lorentzian continuum rather than an unstructured or maximally connected one. That demonstration is computational, not textual, and is named here as the theory's next required step rather than attempted in this paper.

Level I. The Ontological Claim

Every mature physical theory eventually meets a question it did not start with: why does anything change at all. Classical mechanics, quantum theory, and general relativity all answer this by assuming change from the outset, folding it into a time variable that is simply given as part of the theory's starting furniture. This paper refuses that shortcut. It asks what minimal condition must hold before anything can be said to differ from anything else, and proposes that this condition, distinguishability, is not a property objects happen to have but the generative source from which objects are built.

Most of the philosophical tradition treats being as prior to becoming. A thing exists, and only afterward does it happen to change. This paper inverts that order. It proposes that becoming is prior to being, and that a stable thing is a becoming that has, for as long as it lasts, stopped producing new distinctions faster than it can absorb them. An object is not a thing that happens to move through time. An object is a locally slowed rate of differentiation, a place where change has become self similar enough to be given a name.

Under this reading, energy is not a property that things possess. Energy is the rate at which the world is still becoming, measured wherever one looks. Time is not a stage on which becoming occurs. Time is what becoming looks like once it has accumulated long enough to be compared against itself. Change is not something added to an otherwise static universe. Change is the only thing that was ever there. A perfectly frozen instant is not a simplified description of reality. It is a description of nothing, since nothing distinguishable can exist without the activity that produces distinguishability in the first place. Stillness is not the baseline from which change departs. Stillness is a limit that change approaches but never reaches.

What distinguishes this proposal from a general relational ontology, and from causal set theory in particular, is a specific chain of claims rather than a single primitive. Causal set theory builds spacetime directly from order [1]. This paper proposes that energy and time arise first, from the same order, before spacetime is reached: order gives rise to an intrinsic activity, activity gives rise to energy and time together, and geometry and matter follow only afterward, as consequences of energy and time rather than as co equal companions to order. That specific claim, not the existence of a partial order, is what this paper must ultimately defend, and it remains a proposal rather than a demonstrated result until the computational work described below is carried out.

Level II. Mathematical Dynamics

The Differential Structure

The formal object of the theory is a differential structure, a tuple D consisting of a set V of generated poles, a precedence relation among them, a differentiation operator, and an intrinsic measure. The precedence relation must be a strict partial order: irreflexive, so that no pole precedes itself, and transitive, so that precedence composes.

Admissibility

A transition from a structure G to a structure G prime is admissible if and only if G prime is obtained from G by introducing exactly one new pole p, together with a set R of new precedence relations between p and elements already present in G, subject to two conditions. Closure: after adding the relations in R and taking their transitive closure, the resulting relation must remain irreflexive. Locality: every relation in R must directly involve p, so that the new pole cannot be used to rewrite relations among poles that already existed.

Given the locality condition, the only way to introduce anything with positive intrinsic weight is to introduce a new element, since relabeling existing poles carries no new weight under the invariance result proven below, and modifying existing relations directly is already forbidden by locality. Single pole creation is therefore the minimal act consistent with locality that can carry positive weight. This is a claim about what follows from the axioms adopted here, not a claim about what nature is ontologically compelled to do. The theory states that pole creation is necessary given locality and positivity. It does not yet claim that locality itself is a metaphysical necessity rather than a modeling choice, and that distinction should not be allowed to blur.

Composition

Two differentiation events D one, taking G to G prime, and D two, taking G prime to G double prime, compose to a single history taking G directly to G double prime. Composition is associative, since it is sequential application of admissible transitions. The identity transition, which introduces no pole and changes nothing, is included for algebraic completeness, so that differentiation histories form a monoid. The identity transition is an algebraic device, not a physical differentiation event. Level I's claim that a perfectly frozen instant describes nothing distinguishable is not contradicted by this: the identity element makes the composition law well behaved as mathematics, but the theory does not claim the world can pause.

Combinatorial Invariance Under Order Isomorphism

Call two differential structures G and G star equivalent if there exists a bijection between their pole sets that preserves precedence in both directions, an order isomorphism.

Proposition. If G and G star are equivalent under an order isomorphism phi, and D is an admissible differentiation event on G introducing pole p with new relation set R, then the corresponding event D star on G star is also admissible, and its intrinsic weight equals the intrinsic weight of D.

Proof. Admissibility is preserved because phi is an order isomorphism, so the closure and locality conditions transfer directly to their images under phi. The intrinsic weight of D is the count of new precedence relations created between p and existing poles. Since phi is a bijection, it carries R to a set of the same cardinality, and these are exactly the new relations created by D star. A bijection preserves cardinality, so the two weights are equal.

This is an elementary result, and it should be described as one. It shows invariance under relabeling and isomorphism. It does not show, and this paper does not claim, invariance under coarse graining, invariance across distinct differentiation histories that reach the same structure by different routes, or Lorentz invariance, each of which is a separate and strictly stronger requirement. The weight w and the measure mu are, as of this proposition, combinatorially intrinsic: independent of labels and of representation. Nothing beyond that should be read into the term.

Toward a Transition Law: A Minimal Candidate

Admissibility constrains what can happen. It does not say which admissible transition occurs, and this remains the theory's largest structural gap. This paper states one concrete candidate rule, labels it explicitly as a candidate rather than a derivation, and computes its consequences on a minimal case. Throughout this section and the remainder of the paper, the rule is referred to as a candidate dynamics, not as the theory's dynamics, because the primitive ontology of Level I and the formal skeleton of this Level do not depend on this particular rule being correct. If the candidate fails, that failure falls on the candidate, not on differentiation as a primitive.

The candidate rule: given a structure G with admissible differentiation events D one through D k, the probability of each is proportional to the exponential of its intrinsic weight. This assigns strictly positive probability to every admissible transition, including one that adds no new relations, and it favors transitions that produce more relational structure in a graded rather than absolute way.

After a first pole has been generated, two admissible continuations exist for the next differentiation event: a second pole with no relation to the first, weight zero, or a second pole with one relation to the first, weight one. Under the candidate rule, the probability of the unrelated continuation is approximately zero point two seven, and the probability of the related continuation is approximately zero point seven three.

Two problems with this candidate must be stated without softening, since naming them is more valuable than letting the rule stand unchallenged. First, the specific exponential form was chosen for definiteness and to avoid zero probabilities, not derived from any deeper requirement; other forms remain equally available, including one that weighs generation against an as yet undefined coherence or persistence cost, a direction this paper names but does not develop. Second, and more seriously, a rule that always favors more relational structure, applied repeatedly, tends toward maximal connectivity, which is not obviously compatible with a universe that produces sparse, low dimensional, structured geometry rather than a densely interconnected one. This tension is not hidden here because it is, at present, the theory's most informative feature. A primitive can be well posed while its candidate dynamics is wrong, and a transition law specific enough to fail this clearly is more valuable than no transition law at all, since it can now be tested and discarded rather than remaining permanently unfalsifiable.

Level III. Physics

Energy and Extensivity

Physical energy is extensive: the energy of two independent, non interacting systems together equals the sum of their individual energies. The intrinsic measure mu is additive over independent regions by construction. An earlier treatment of activity divided each chain's accumulated weight by that chain's own step count, and the ratio of two extensive quantities is not itself extensive, so combining independent chains of different lengths did not make their activities add correctly.

The repair measures duration by a shared reference rather than by each chain's own steps. Fix a differential clock, as defined below, and let T denote the reference duration it assigns to the interval under consideration. Define physical energy as E equals kappa times accumulated weight divided by T, with T shared across every subsystem being compared. Two independent chains observed over the same reference interval then have energies that sum correctly, because their weights already sum correctly by the additivity of mu, and both are divided by the same T.

This repair also bears on why energy should be linear in accumulated weight rather than some other power, though the argument should be stated with its assumptions explicit rather than compressed. Suppose energy is some function f of the accumulated weight, E equals f of mu, and suppose f satisfies the additivity condition f of mu one plus mu two equals f of mu one plus f of mu two for independent regions, since this is exactly what extensivity requires. This is a Cauchy functional equation. Under a mild regularity condition, for instance that f is monotonic or merely bounded on some interval, conditions physical energy should reasonably satisfy, the only solutions are of the form f of mu equals kappa times mu for some constant kappa. Without some such regularity condition the Cauchy equation admits pathological, nowhere continuous solutions that no physical quantity should be expected to follow. Stated this way, the conclusion is that linearity is forced by extensivity together with an ordinary regularity assumption, not by extensivity alone, and that qualification should not be dropped.

It should be stated plainly what has and has not been shown. What has been shown is conditional: if the intrinsic measure is to be related to physical energy at all, and if that relation is required to be extensive, then under a mild regularity condition the relation must be linear. What has not been shown is that the intrinsic measure is in fact related to physical energy. The constant kappa remains free, undetermined by anything internal to this theory, and fixing it requires an independent physical comparison that has not yet been made.

Time

A chain of differentiation events supplies order but not duration. A differential clock is a persistent pattern whose intrinsic activity, computed under the definition above relative to some provisional reference, remains approximately constant across successive segments of its own unfolding. Given such a clock, the duration of another process is the number of times the clock's regular segment must repeat to remain comparable, in accumulated weight, with that process. This still relies on counting repetitions, a primitive act of comparison rather than something derived from nothing, and that reliance is acknowledged rather than hidden, since comparison is required for the theory's founding claim, that two things must be distinguishable before anything can be said about them, to make sense at all.

This construction depends on the existence of persistent, regular differential patterns, and that existence has not itself been established. Whether the candidate dynamics of Level II, or any dynamics, reliably produces patterns regular enough to serve as clocks is an open dynamical question, not a definitional one. Time, in this theory, is not fully derived until persistent regular differential patterns are shown to exist dynamically, rather than merely defined as what a clock would be if one existed. This is a bootstrapping dependency rather than a circularity: the definition of time is not self referential, but it is not yet self sufficient either, since it borrows an existence claim from a part of the theory, the dynamics, that remains a candidate rather than a result.

Toward Lorentz Invariance

This remains the central physics target the theory has not met. The relevant existing result, due to Bombelli, Lee, Meyer, and Sorkin [1], is that a partial order generated by random sprinkling into Minkowski spacetime with density fixed by a Lorentz invariant Poisson process favors no preferred frame on average, even though each individual realization is discrete. This shows that the naive objection to any discrete foundation, that discreteness automatically breaks Lorentz invariance, does not apply to every discrete structure, and in particular need not apply to a locally finite partial order of the kind considered here.

What follows from this is only that the objection is not automatically fatal to a partial order of this general type. It does not follow that the transition law candidate proposed above, or any other specific rule this theory might adopt, generates structures whose statistics match a Lorentz invariant sprinkling. Demonstrating that would require showing that the differentiation process, in its continuum limit, produces a distribution over structures statistically equivalent to a Poisson process of fixed density in Minkowski spacetime. This has not been attempted here, and the citation above should not be read as progress toward it, only as evidence that the target is not ruled out in principle.

Geometry and Matter

Persistent relational connectivity, once sufficiently regular and, if the Lorentz question above is resolved, statistically compatible with a Lorentzian continuum, is proposed to admit an effective notion of adjacency, distance, and curvature. No derivation of a metric or curvature term from the differential structure is given in this paper.

Matter is proposed, not established, as an attracting equivalence class of differential patterns under the admissible dynamics: a class such that differentiation histories beginning near a representative, meaning histories agreeing with it up to a bounded number of differing admissible choices, tend to return to representatives of the same class rather than drifting arbitrarily far under continued differentiation. This is more precise than describing matter as simply a stable pattern, because it specifies what stability would have to mean in terms of the dynamics. No example in this paper demonstrates such an attractor, and none could, since no structure considered so far is large enough for attracting behavior to be meaningful, and since the dynamics remains a candidate rather than a settled rule. The definition is offered so that a future demonstration knows precisely what it must show, not as a claim that one has been shown.

Relation to Existing Work, and What Is Actually New

The admissibility rule is, deliberately, close to the definition of a causal set as a locally finite strict partial order, and the appeal to the Bombelli, Lee, Meyer, and Sorkin sprinkling result [1] is a direct borrowing, named as such. The thermal time proposal of Connes and Rovelli [2] remains the closest existing treatment of time as extracted from a system's own statistics, and the clock construction here is a combinatorial cousin of that idea.

Given this proximity, the honest answer to what is genuinely new here is a specific chain of claims, not a new primitive object. Causal set theory proposes that order gives rise to spacetime. This paper proposes a longer chain: order gives rise to an intrinsic, combinatorially invariant activity; that activity gives rise to energy and time together, with energy's functional form now derived from extensivity under stated assumptions rather than asserted by analogy; and only afterward, and only as an unproven target rather than an established result, does the same structure give rise to geometry and matter. That energy and time are proposed as the first emergent pair, prior to a well posed notion of spacetime rather than a by product of spacetime once it exists, is the paper's central thesis. Every other construction in this paper, the admissibility rule, the transition law candidate, the attractor definition of matter, is scaffolding built to make that thesis checkable rather than merely stated.

Epistemic Status

Established by an elementary but genuine argument: the admissibility rule is precise and checkable; the invariance proposition is proven for order isomorphic structures in general, and holds directly under any explicit relabeling; and the energy functional is shown to be forced to linear form by extensivity together with a stated regularity condition, rather than assumed by analogy.

Offered as an explicit, labeled candidate rather than a derivation: the exponential transition law, whose correctness relative to alternative candidates, including forms that weigh generation against a coherence or persistence cost, is unknown; and the attractor definition of matter, which specifies what would need to be shown without showing it.

Named as open and unresolved, without qualification: whether the candidate transition law, or any variant of it, produces a continuum limit at all; whether that continuum limit, if it exists, is statistically Lorentz invariant in the sense made precise by the sprinkling literature; whether persistent regular patterns adequate to serve as clocks actually arise under the candidate dynamics; and whether curvature and an Einstein type equation follow from the intrinsic measure in that limit. The candidate rule's evident tendency toward maximal connectivity under repeated application is flagged as a specific, likely failure mode rather than left implicit, since a theory whose first concrete dynamical candidate can be shown to fail in a specific, nameable way is in a stronger position than a theory with no dynamical candidate to test at all. Should this candidate fail under computation, the finding falls on the candidate, not on the ontological claim of Level I or the formal skeleton of Level II.

Final 

This paper has proposed that the primitive fact of the universe is an act, differentiation, from which energy and time are proposed to arise together, prior to geometry and matter, as two faces of one intrinsic activity. It has corrected a real defect in the earlier treatment of energy, deriving linearity from extensivity under an explicit regularity assumption rather than asserting it. It has stated a concrete transition law candidate, evaluated it on a minimal case, and named its most serious weakness, a tendency toward maximal connectivity, as a specific and useful way for the candidate to be shown wrong rather than as a difficulty to be minimized. It has sharpened the justification for single pole creation while keeping the distinction between axiomatic necessity and ontological necessity intact, and has separated the algebraic identity of the composition law from the physical claim that nothing genuinely static occurs.

Every mathematical result offered here is elementary, and none of them should be read as more than what it is. The paper has not shown that a continuum limit exists, has not shown Lorentz invariance, has not shown that persistent clocks arise, and has not derived an Einstein equation or any measured quantity of physics. These cannot be completed by further writing. The next step is computational: generate large differential structures under the candidate transition law, or under variants of it, and examine their connectivity, their degree distribution, their approximate dimension, and whatever causal structure they exhibit, to see whether the candidate's predicted drift toward maximal connectivity is as fatal in practice as it appears in principle, and whether any nearby variant avoids it while remaining as simply stated.

Until that work is carried out, what stands here is a philosophically serious proposal with a formal skeleton precise enough to state its own open problems exactly, and a first, explicitly falsifiable dynamical candidate, clearly separated from the primitive it was built to test. It should be judged as exactly that, and no further, until the computation is done.

References

[1] Bombelli, L., Lee, J., Meyer, D., & Sorkin, R. D. (1987). Space-Time as a Causal Set. Physical Review Letters, 59, 521–524.

[2] Connes, A., & Rovelli, C. (1994). Von Neumann Algebra Automorphisms and Time-Thermodynamics Relation in Generally Covariant Quantum Theories. Classical and Quantum Gravity, 11, 2899–2917.