Pre-Closure Field Distribution: An Ontology of Quantum Field Theory and Its Interface with the Four Forces
前闭合场分布——量子场论与四力接口的本体

Part I. The Ontology of Pre-Closure Field Distribution

— the field is not a continuous substance but a distribution of pre-closure capacity over discrete cells; the continuous field is only one of its faces at the coarse-grained layer

1. Starting point

The preceding papers were framed, by and large, around a single quantum, a single closure, a single menu. The Born-rule paper [1] gave the weights; the measurement paper [2] gave the single content that is read out; the decoherence paper [3] placed onto the causal slot that objective menu (menu-information) which, after decoherence and before settlement, is dephased yet still many-holding. But these still concern the pre-closure capacity of one object — one particle, one system, one superposition. Nature, however, does not take the single quantum as its norm. Once pre-closure capacity is no longer tied to a single object but is spread across many cells, many modes, many closure-addressable locations, the single-particle language recedes into an effective approximation and the language of field theory appears. The decoherence paper began the turn from the microscopic to the macroscopic; the present paper is its sequel, and asks: when that menu no longer belongs to a single quantum but to an entire collection of cells, what is it?

The scope of this paper must first be fenced. It does not redo the four-forces series, does not derive the Standard Model, and does not restate the gauge structure; it asks only what the pre-closure "field distribution" is ontologically within this framework, and how it interfaces with the algebra and representations of the four-forces series. The path integral, the full continuum limit, and the emergence of classical mechanics are all left to the concluding paper of the series. In treating the field operator, creation and annihilation, the vacuum, and virtual particles, the aim is to place them back into the ontology of pre-closure capacity, without altering any of their observable predictions — this framework issues predictions where a divergence is detectable, and only re-reads where it merely reproduces standard physics; the present paper belongs to the latter.

2. Discreteness, and Lorentz

The first step on this road requires that one thing be settled in advance. This framework has a plain, hard prior: every physical quantity, at a sufficiently small scale, is discrete (after the foundations paper [4]; the decoherence paper [3] has already affirmed this for fields and for probability clouds). This discreteness is not a dispensable categorical abstraction — it has a real substrate. The relativity papers of this series [5] long ago established the absoluteness of the Planck scale: c = l_P/t_P is the universal pass-through rate, and the Planck scale l_P stands as the real, absolute natural scale — invariant, unaltered by the system, and unaltered by the distortion of mass. A distinction must be drawn: what is invariant and absolute here is the graduation of that ruler (the reality of the scale l_P) — l_P is a universal constant, its value the same for all observers, just as c is; this does not mean that the spatial extent of a given cell in a given frame is frame-independent. The true per-cell Lorentz invariant is the cell's causal capacity and 4DD super-volume (see the next section); a cell's spatial extent, as will be seen below, contracts with the system. Whether the concrete physical scale of this discreteness is exactly l_P, and how it descends to the field layer, is a question for physics to measure; this paper takes the Planck scale only as the real, absolute natural candidate anchor (after [5]) and does not cash it out, on the physicists' behalf, into any definite value — the specific discrete scale awaits experiment (after [3], [4]).

Discreteness has always sat uneasily with Lorentz invariance — a grid with an absolute scale, laid out in space, seems to carry a preferred frame, so that under a Lorentz transformation the grid shows its hand; yet the Lorentz invariance of quantum field theory has been tested to extraordinary precision, with no crack found to date. A framework that reads the field as something over discrete cells must therefore first answer this: how does the real, absolute Planck discreteness coexist with Lorentz invariance?

This question has been addressed in the relativity papers of the series [5], and is inherited here rather than re-established. Its essence lies in a two-layer structure. Underneath is the Planck substrate, absolute and universal — light, as a 2DD wave, carries no information and travels at this layer, which is why its speed is the same for all observers; above it is the causal slot, whose scale varies with gravity and with motion. A relativity paper [5] makes this plain: a cell's spatial length varies with the system — contracting to 1/γ along the direction of motion — while what Lorentz holds invariant is not that length but the cell's 4DD super-volume (R₀³·c·t_P; or, in c = 1 units, R₀³·t_P) and the one bit of causal capacity it carries per cell. In short: what is invariant is the capacity, the super-volume; what is variable is the length. This invariance of the super-volume is recorded, in the relativity paper, soberly as the ontological restatement of the invariance of the special-relativistic volume element (unit Jacobian) — Lorentz is inherited, not derived afresh from a prior; and it is inherited here likewise. Two further points are key. First, that absolute Planck substrate is a layer of ontological substrate, not an empirically distinguishable preferred frame: all inertial observers are treated alike under Lorentz transformation, and to say a cell "truly contracts at the absolute Planck scale" is only an ontological turn of phrase, whose observable consequences are identical to standard special relativity down to the last decimal. Second, this substrate is not a rigid square lattice with preferred directions either — were it a square grid, the anisotropy of its diagonals would break the rotations and boosts of Lorentz; it is rather a relational network whose nodes are that conserved causal capacity, and precisely because what it conserves is capacity (super-volume) rather than length along some direction, it can, at the coarse-grained effective layer, accommodate the continuous symmetry of Lorentz. So the real, absolute Planck discreteness and Lorentz invariance run in parallel without conflict.

Here two genuinely distinct concepts must be kept apart, lest they encumber one another. The Planck scale speaks of the real, absolute substrate scale; the causal slot is a categorical threshold — an aggregate of substrate that crosses the threshold of causal closure and becomes a carrier of causal information is what is called a causal slot (after the information paper of this series [6]). The former is absolute and invariant, speaking to the reality of scale; the latter varies in size with conditions, speaking to the crossing of a category. Each serves its own sense: discreteness has its real Planck substrate, while the causal slot is a categorical division above that substrate. The earlier thesis that "the causal slot is a category, not a size" speaks precisely to the latter being a category — it does not annul the former's being real; the non-thresholdhood of size and the reality of scale are two separate matters.

As for the continuous coordinate x in the field operator φ(x), this paper places it thus: it is not a private name of a fundamental cell but an effective coordinate chart furnished at the layer of apparatus. The fundamental layer is a closure-addressable cell-complex, not a lattice pre-embedded in some fixed Euclidean space; the continuous x is the chart that this complex acquires when coarse-grained under a given apparatus and spacetime background. Hence x belongs to the effective read-out layer, not the fundamental ontological layer; the covariant transformations among such charts are not derived here, as they fall to the relativity papers of the series. The matter of covariance of settlement in a relativistic setting — that measurement admits no instantaneous collapse across a spacelike interval — falls under the same jurisdiction, and is inherited here in framework without being unfolded.

With the discreteness side now secured — a real, absolute Planck substrate, the causal slot as a categorical division above it, Lorentz invariance carried by that two-layer structure — only then can one ask what, precisely, this pre-closure field distribution is.

3. The load-bearing ontology: discrete cells, local capacity, gluing

Having said that the continuous field is not fundamental, one must say what the fundamental layer is. First a ruler is needed, to adjudicate whether a thing is a description at the effective layer or the ontology itself — and this ruler runs through the whole paper. In this framework, the adjudication rests not on the names "mathematics" and "physics" but on the fact of "continuous" versus "discrete": whatever is continuous and unfathomable abstraction is description at the effective layer; whatever is discrete and real is the load-bearing ontology — and that load-bearing discrete ontology falls precisely at 3DD (after the DD ladder of the relativity series [5]: 3DD is the locus of mass-binding and spatial aggregation). The continuous field, the operator algebra, the propagator and Green's functions, the path-integral measure, and like formal objects are all continuous abstractions, descriptions at the effective layer; the discrete and real cells, the capacity on the cells, and the gluing between cells are the 3DD load-bearing ontology. The matter must be put fairly: the operator algebra is a representation-independent grammar, recording observables, recording the structures of locality and commutation, and carries very real weight; this paper does not disparage it, nor does it take the road of "the continuous operator algebra is the ultimate ontology" — it merely keeps the load-bearer seated in the discrete cell-capacity and gluing rules, with the algebra as their effective read-out form. However precise and representation-independent the continuous algebra may be, in this framework it is the effective grammar of the ontological structure, not the ultimate substrate posited.

The load-bearing ontology is this picture: a sheet of discrete cells — whose substrate is just the real, absolute Planck substrate of the preceding section, still beneath the causal slot — each cell carrying a local, pre-closure ρ-OR capacity, and between adjacent cells a set of gluing rules prescribing how local capacities are to be compared, transported, made coherent, and coupled. To give a precise skeleton: let the set of cells be C = {cᵢ}; on each cell cᵢ there is a local capacity algebra 𝔄ᵢ and a local ρ-OR capacity ρᵢ (this ρ is not the notation of a density matrix but of local capacity), and between adjacent cells a gluing rule Gᵢⱼ. The field distribution is not a set of already-realized contents but precisely these three, {𝔄ᵢ, ρᵢ, Gᵢⱼ}. That gluing between adjacent cells is the origin of what will later, at the four-forces interface, appear as the gauge connection — held here for the section on the four-forces interface.

The continuous field operator φ(x) thereby finds its place in this picture: it is not "a continuous substance truly suffused at the point x," but, when the apparatus coarse-grains this sheet of discrete cells and lays out the continuous coordinate x, an effective read-out issued upon this capacity distribution. The point x is not a fundamental cell; a read-out at a point is the product of pushing the coarse-graining projection to its limit. So the continuous field is one face of this discrete capacity at the effective continuous layer — as an effective construction it is real and effective, neither the fundamental substrate nor a mere computational convenience. As for that set of canonical commutation and anticommutation relations — the algebra of field theory — it is a representation-independent description of this capacity structure: it describes the very ontology of cells and capacity, not some thing beside it.

In this connection one may treat a matter long reputed subtle in field theory — Haag's theorem [20]. Continuous quantum field theory has infinite degrees of freedom, so the canonical commutation relations admit infinitely many unitarily inequivalent representations, and the interaction picture cannot stand mathematically — this is exactly where "representation-independence" is most strained under the continuum idealization. This framework does not claim to have derived or disposed of Haag's theorem from SAE; what Haag exhibits is, less a minor technical ailment that this paper can directly "solve," than a structural warning: do not mistake the effective continuous interaction picture for the fundamental ontology. The discrete prior of this framework makes here only a finite-layer settlement — at the level of a finite-cell regularization, a finite region, and a finite number of canonical degrees of freedom, Stone–von Neumann–type [21] uniqueness of representation can serve as a benign anchor (this is in kind with what any UV regularization does at a finite cutoff; this paper takes it as one corroboration of the discrete prior, not as a boast of unique resolution); as for the continuum limit, the infinite-volume limit, and how the interacting representation is rigorously recovered, these remain with the concluding paper of the series and with technical specialist papers. So this paper takes from it only the support for one ontological judgment — that the continuous field is not the ultimate substrate — and hangs the step of "rigorous recovery" plainly outside the door.

This juncture also marks out, for this paper, a formal boundary of failure (of the C-type below): should that finite-cell capacity network prove unable to yield a sensible continuum inductive limit, or unable, in that limit, to recover the scattering, locality, and renormalized observables of standard field theory, then this whole formalist re-reading fails here. Haag's difficulty is thus not an entry point for outside attack but the most honest falsifier this paper itself avows.

In this connection one may also extend an explicit interface to algebraic quantum field theory (AQFT). AQFT erects a net of local algebras O ↦ 𝔄(O), recording observables, locality, and commutation structure; in this framework that net is not a straw man of "the algebra as ontology" but precisely the observable shadow of the load-bearing cell–capacity substrate over an effective region O — writable as O ↦ CoarseGrain({𝔄ᵢ, ρᵢ, Gᵢⱼ}_{cᵢ⊂O}). So AQFT's local algebras are not overturned in this framework but seated in the effective layer: they are the coarse-grained read-out of the ontology, not the ontology itself. This interface is precisely an invitation to those who work in algebraic field theory to enter the categories of this framework.

Here the paper's three ontological tiers are stated once, to spare scattered reading. First, the fundamental and load-bearing: the discrete cells, the capacity on the cells, the gluing between cells — the 3DD real substrate (fundamental). Second, the real-but-effective: the continuous field (as an effective construction), the L3 "particle," and the operator algebra (as the representation-independent grammar of that load-bearing ontology, an effective description) — all real but not fundamental (real, not fundamental): they are indeed real things, yet not the ultimate substrate, but emergences of that substrate at the effective continuous layer, or its grammar. Third, the purely representational artifact: the picture of virtual particles and off-shell internal lines — a computational thing (artifact), with no ontological name (see Part III). One confusion must be marked: the continuous field (or operator algebra), as "an effective construction (grammar)," is real and effective; as "a continuous mathematical object, or something taken to be the ultimate substrate," it is merely a description at the effective layer and is itself not the ultimate substrate — the two senses must be parted, and the reality of the former must not lead one to take the latter for the ontology.

4. Field operators, Fock, and creation–annihilation: the capacity-layer reading, and what a "particle" is

Following the preceding section, the whole ornate apparatus of second quantization reads, in this framework, as falling back onto the capacity layer.

What the field operator asks is: "if a closure-addressing is set here, at this mode, which capacities can be recombined and read out?" — it is a query, not a lump of substance. The "distribution" in operator-valued distribution thereby finds its footing: because the point is not a fundamental cell, taking a value at a point is the manner of speaking when the coarse-grained apparatus projection is pushed to its limit.

The creation and annihilation operators are then neither making nor unmaking things. a†|0⟩ does not, ontologically, conjure a little ball out of nothing; it raises a certain mode in the vacuum baseline to a single-excitation capacity configuration — a local recombination on the topological features of the 1DD-to-3DD layers; only when it undergoes that settlement with a detector and a record does it become a single content. The annihilation operator likewise: it is a recombination of this capacity ledger, not the annihilation of a remainder — the remainder cannot be annihilated, a law consistent throughout this framework, and so here too. Fock space is then not a warehouse stacked with little particles but a capacity menu of occupation numbers |n₁, n₂, …⟩ — a menu of closure-readable occupation-number configurations. And the scattering matrix S evolves precisely the unitary flow among these menus, a matter at the capacity layer; settlement is not within it — quantum field theory itself only evolves the menu and does not pick out the single content; what picks out the content is measurement, the settlement of detection.

Here a categorical misidentification readily arises and must be guarded against in advance. Bosonic multiple occupation and fermionic single occupation look temptingly like "OR" and "AND" — bosonic multi-occupiability resembling ρ-OR, fermionic single-occupation (the exclusion principle) resembling ρ-AND — yet this mistakes two axes for one. Statistics — commutation and anticommutation, the vanishing of a†a†, the single-occupation of exclusion — is kinematics at the capacity layer: it lies across the capacity layer and prescribes how much each mode can hold. ρ-OR and ρ-AND are the vertical axis of closure: the many-holding before closure, and the singleness after settlement. Even in a fermionic field, within each candidate content a given fermionic mode is still pinned by the exclusion principle; the many-holding that ρ-OR means is the multiplicity of modes (modal plurality), not that each mode may be occupied without limit. The two must not be confounded — to confound them is to be misled into thinking that this framework's ρ-OR is innately biased toward bosons, which is wrong.

What, then, is a "particle," given that the field operator makes nothing and Fock is but a menu? It is not a fundamental object. A particle is a capacity pattern of the field which, when it is sufficient to be recorded, presents as a real excitation that is on-shell and closure-ready — already addressed, indeed present in the field, yet not yet constituting that single content. In the language of this framework's magnitude ladder [4], a particle is an L3 "mark" and not an L3 "constitution": a real marking (real), not a basic constitution (fundamental), and not yet the content born at settlement. Quantum field theory dissolves the particle entirely into the field operator and no longer takes it as fundamental; this framework does not dissolve it but grants it a real station, L3-marked-not-constituted — real, not fundamental. This adjudication looks fine-grained, yet is the linchpin of what follows: in treating virtual particles, it is precisely for the lack of such addressing and such a real station that one judges them to have no ontological name; in treating antiparticles, it is precisely because they are a real excitation of the opposite side, and not some other thing.

5. The menu lifted onto the field: from a single object's menu to a field-level event menu

The menu-information established in the decoherence paper [3] — the answerability-before-the-question, capacity rather than content — was, in a single system, no more than up and down, spin plus and minus, path A and path B, this position-interval and that. But once lifted onto the field, this menu grows by a level: it is no longer "where is some particle" but a candidate set of whole field configurations, or of occupation configurations.

Take a collision as an example. Before settlement, the field-layer menu asks not "where is some particle" but the various possible event topologies: two photons, an electron–positron pair, several jets, or a channel that has lost energy; decoherence and detector coupling make these event topologies a menu on the slot — dephased, yet still many-holding; while the real content is, in the record of a particular detection, which set of cells truly clicked, how much energy was deposited, and how the track joined up into a line.

So the menu-information in the field is not "a single object's classical menu" but "a field-level event menu." This is to extend the menu of the decoherence paper from measurement to the field, not to restate what it has already said.

6. The vacuum, and zero-point energy

The vacuum, then, is not nothing. It is that baseline ρ-OR capacity when there is no settled excitation whatever — every cell across the sheet in its pre-closure ground state, maintaining mutual connectedness, not yet engaged by any local settlement. It is not even an absolutely singular object: it is the "no settled excitation" relative to some grammar of closure-addressing, not that absolute nothing indifferent to grammar — this dimension of relativity will show its weight in field theory on curved spacetime, held here for the relativity and cosmology papers.

Zero-point energy, then, is not a trick of virtual particles borrowed and returned. The ineliminable thing must be stated precisely: what is ineliminable is the nontrivial capacity structure and correlation of the vacuum baseline, not necessarily the absolute numerical value of the zero-point energy under some mode decomposition — that absolute value can be handled by normal ordering or renormalization, a matter of the effective formalism, and does not mean that the vacuum baseline is ontologically empty. The zero-point energy is, under a given mode-grammar, a signature of the vacuum baseline. The phrasing must remain exact here: the vacuum is a zero-occupation state, so this baseline is "the ineliminable minimal capacity structure," not "many-holding" — many-holding is a matter elsewhere; the vacuum is precisely that lowest, yet still nonzero, capacity floor.

A firewall must be hung at the close. This capacity baseline and the 4DD content-bearing source — that to which gravity responds — must be treated as two layers. To sum the zero-point energies of the modes naively, and take that, without categorical scrutiny, directly as the content-bearing source, is improper; this framework separates the capacity baseline from the 4DD read-out into two layers, and so warns that this step must not be taken rashly. Hence the Casimir effect [25], the Lamb shift [26], and the anomalous magnetic moment [27] are posterior anchors of the structure of the vacuum, not grounds for "the absolute numerical value of zero-point energy gravitating directly." As for the question of the magnitude of the cosmological constant itself, that is for the cosmology and gravity papers to handle, and is outside this paper's jurisdiction.

7. Entanglement entropy, and the area law: cross-boundary capacity-correlation

The preceding sections set the field's ontology on those three, {𝔄ᵢ, ρᵢ, Gᵢⱼ}; that being so, the entropy of entanglement, and that law of area long reputed profound, acquire an ontological reading in this picture too — and this reading is exactly where the framework is strong.

First two genuinely distinct matters must be parted, lest they encumber the earlier paper. Entanglement as a phenomenon — the indecomposable ledger, the ledger of A and B together not equal to the direct sum of their separate ledgers (𝔏_AB ≠ 𝔏_A ⊕ 𝔏_B) — has been established in the entanglement paper [16], and is not redone here; this section asks only what, ontologically, is the entropy of entanglement between region and region, in the vacuum, in field theory.

Its ontology is, once again, in the gluing. Take a region A and its exterior Ā; the entanglement between the capacity within A and that within Ā is borne by those gluings Gᵢⱼ that cross the boundary ∂A — the bundle of capacity-correlations between the cells of A hugging the boundary and the cells of Ā hugging the boundary. Reducing to A, taking a trace, and computing its entropy (written conventionally S_A, with reduced operator ρ_A; this ρ_A is the standard reduced density matrix, and is none of the earlier local-capacity ρᵢ, nor ρ-OR — do not conflate them) measures precisely this bundle of gluings cut by the boundary. (A word, further, to carry the previous section: since §2 holds that a cell's spatial extent contracts with the system, here too "the number of boundary cells" and "the area" transform together with the system, and the S_A measured is a covariant quantity; the area law does not covertly assume that cell extent is frame-independent.)

From this, that law of area — the leading term of entanglement entropy proportional to the area of the boundary, not the volume of the region — acquires, in this framework, a geometric reading (graded here as a T2 ontological reinterpretation, not a universal theorem). Its rationale must be stated with two premises together: first, that gluing is local, only between adjacent cells; second, that what is at issue is the vacuum baseline (or a low-excitation state). Under these two premises, the principal channel of entanglement between A and Ā is the layer of cells hugging the boundary, and the number of boundary cells is proportional to the area of ∂A, not the volume of A. It must be stated plainly: by "local gluing" alone, graph-theoretically, one obtains only an upper bound on the entanglement (a min-cut bound); for a highly excited or thermal state, even on a locally glued network, the entanglement entropy can obey a volume law. So the area law shows itself because local gluing is conjoined with the low-excitation topology of the vacuum baseline, not because local gluing forces it single-handed. This bound can be marked quantitatively: writing χᵢⱼ for the effective capacity-dimension of the cross-boundary gluing Gᵢⱼ, the leading entanglement borne by local cross-boundary gluing satisfies S_A ≤ Σ_{⟨i,j⟩∈∂A} log χᵢⱼ; and if the boundary gluings are broadly homogeneous (χᵢⱼ ≈ χ) and near-saturated, the leading term ~ |∂A|·log χ, proportional to the number of boundary cells. The S ~ area/ε^{d−2} that standard field theory obtains with some labor (the entanglement entropy of a free field, computed by Srednicki [18]) is, in this framework, a direct shadow of the locality of gluing (on the vacuum baseline). Its reach must also be bounded: critical systems (e.g. the logarithmic scaling in 1+1 dimensions), area-law violation due to a Fermi surface, gauge-field edge modes, topological entanglement entropy, and corrections from long-range interactions all hinge on specific field-theoretic and many-body conditions, and are not adjudicated in this section — this section establishes only the leading area term for the vacuum, local, short-range case, as a T2 reading.

And the UV divergence within that S_A — the infinity as ε → 0 — is again a developing-out, of one piece with the section on the ultraviolet (Part III): it is just "the number of boundary cells → ∞" in the continuum limit; the discrete cell gives it a natural floor. Entanglement entropy is precisely the cleanest UV-divergent quantity in field theory; this framework marks only its developing-out here and does not compute its coefficient (nor treat the logarithmic corrections that massless fields carry in low dimensions), in the same discipline as the renormalization section.

This section also brings the vacuum section down to the concrete. §6 said the vacuum "maintains mutual connectedness"; that connectedness, at the boundary of a region, is precisely the entanglement entropy of the vacuum. The vacuum is not nothing, and the capacity structure of its baseline is real — this sentence acquires, at entanglement entropy, a concrete form: the vacuum's entanglement entropy measures that gluing which crosses the boundary and sustains the connectedness of the two sides.

A firewall must be hung at the close. Entanglement entropy proportional to area — black-hole entropy S = A/4 likewise proportional to area — is a tempting resonance. In this framework the area law of entanglement entropy comes from the locality of gluing (on the vacuum baseline); as for the entropy produced by that special boundary the black-hole horizon, the Page curve, the structure of islands, and their concordances and discordances with the standard replica-wormhole calculation, these fall to the paper on black-hole information [17] and the relativity papers — the standard benchmark of holographic entanglement entropy (Ryu–Takayanagi [19]) is here, too, explicitly drawn outside this paper, and this section does not enter its jurisdiction. This paper treats only the entanglement entropy and area law of flat space and a general region, and stops there: it does not derive S = A/4, does not compute the Page curve, and does not touch islands or replica wormholes. The concordances and discordances with the standard island picture, drawn out by the fine-grained entanglement of black holes, are an open problem left for the future (should a quantum-information series be brought into contact with the black-hole-information paper); this paper sets the boundary here and does not enter.

---

Part II. The Relativity Interface and the Four-Forces Interface

— an antiparticle is a real excitation of the opposite side, not the remainder discarded at settlement; the four forces are the grammar by which capacity distributions are glued

1. The antiparticle: a real excitation of the opposite-side 4DD

What quantum field theory genuinely adds, beyond the preceding papers, is relativity. With relativity comes the antiparticle — the positron, the antiproton, and their kind, each with its posterior reality, detectable and recordable. This framework has a ready-made ontological exit, and that exit was not improvised for field theory but was long ago laid down in the cosmology paper of this series [7]: the symmetry of 3DD gives rise to a pair of 4DDs with opposite time-arrows; our side and the opposite side exchange parameters with arrows reversed. The antiparticle dwells on that opposite side — it is a real, on-shell excitation in the conjugate sector of the opposite-arrow 4DD, which, read out by the apparatus on our side, presents as a charge-conjugate, positive-energy content.

Yet the antiparticle is not "the same particle truly running backwards in our side's time." One may, by way of analogy, borrow the old Feynman–Stückelberg [13] reading of "an antiparticle as a particle propagating backward in time" — it is indeed a useful figure — but the ontology of this framework lies not in "a stretch of film run backward in our side's time" but in the manifestation, in our side's read-out, of the conjugate sector of the opposite-arrow 4DD. Pair production, then, is not the conjuring of two substances out of nothing: under coupling to a strong external field or background, a pair of conjugate capacity sectors within the vacuum baseline is raised to closure-addressable excitations, read on our side as one particle plus one antiparticle; their conservations — charge, energy, momentum — are borne by the constraints of the representation layer. Schwinger's strong-field pair production [24] serves as an example, yet it is a standard phenomenon — re-read here, with no new prediction issued.

There is here a concordance worth stating outright, though its weight must be calibrated. The structure of the opposite-arrow 4DD was erected, in the first place, for the time-arrow of cosmology, for the priors of that series [7], and not for particle physics. This paper does not use it to replace standard QFT's derivation of the antiparticle (which requires a full relativistic field representation); what it furnishes is an ontological station — every charge-conjugate sector that standard QFT already provides falls naturally, in this framework, onto the conjugate sector of the opposite-arrow 4DD. So the reality of the positron (a posterior fact already secured in 1932) is, in this paper, not re-predicted but meets, in consilience, an independently established cosmological structure. An a priori structure erected for the cosmological arrow and a posterior fact of particle physics meet and accord here — this is not a new prediction but a nontrivial concordance between a prior and a posterior, marked as such and not vaunted as prediction. A boundary must be added: "antiparticle" here means strictly the charge-conjugate representation sector; for self-conjugate representations (the photon, the π⁰, and the Majorana neutrino were it to exist), the conjugate sector may coincide with the read-out on our side and does not additionally require an empirically distinct second particle.

From this structure follows a falsifiable engagement, though its direction must be calibrated: the reading of the opposite-arrow 4DD requires that every sector which standard QFT can confirm as a particle–antiparticle conjugate pair must be able, in this framework, to settle as a conjugate pairing on the opposite side; should this reading prove unable to accommodate charge conjugation, equality of mass, positive-energy read-out, and pair production, then this interface fails. What is at stake is whether the opposite-arrow 4DD interface can carry the load (after the cosmology paper [7]; not newly fabricated here, and not an independent prediction of mass equality by this paper). The falsifiable point of engagement is precisely here.

This identification must be fenced from three matters. First, that CPT is a theorem, and the derivation of why a particle and its antiparticle have strictly equal mass, belong to the four-forces and relativity series; this paper does not derive them — what it gives is the ontological ground (why CPT is reasonable), not its derivation. Second, why matter and antimatter are asymmetric, why baryogenesis occurs, why our side is matter-dominated, belong to the cosmology series (whence the opposite-arrow 4DD), and are not reached here. Third, the antiparticle is not the remainder of settlement, treated in the next section.

2. The antiparticle is not the remainder of settlement

The decoherence paper [3] established that settlement picks out the single content on the spot, while the unselected terms are not annihilated but enter, as remainder, the 4DD broadcast conservation ledger. A reader may readily form a dangerous misidentification here: since the antiparticle dwells on that opposite side, and what settlement discards is also the side under the reversed arrow, are the two not the same exit?

They are not. The two stand at the same geometric station of the opposite-arrow 4DD, yet are two entirely different things at that station. The antiparticle is a real excitation in the opposite-side sector, detectable by our side — a kinematic actuality, writable into Fock, makeable into a menu, readable by settlement. The remainder of settlement is the destination of those capacities not realized in a closure event — a bookkeeping residue, which travels, along the channel established in the two information papers [10][11], to the 4DD broadcast conservation ledger (its energetic aspect via E=Ic³, established in the mass series [9]). One is the ontology of a sector; the other is the bookkeeping residue of an event; one is detectable; the other is merely recorded on that global conservation ledger and emits no distinguishable gravitational wave.

To conflate the two yields two catastrophic misreadings: first, supposing that the unselected branch in each measurement turns into an antiparticle — then a single Stern–Gerlach experiment would spray out a shower of positrons, absurd in the extreme; second, conversely, demoting the antiparticle to the dross of "unrealized possibility" and refusing to recognize it as a detectable real thing — a disrespect to the posterior fact of the positron. So the two must be kept distinct: the antiparticle is a real excitation of the opposite side (kinematics); the remainder of settlement is the bookkeeping residue squeezed out by one-way-ization (dynamics, headed to broadcast); at the same geometry of the opposite side, they are different things.

3. The ontological seat of CPT, not the derivation of a theorem

This paper does not derive the CPT theorem — that is the physicists' work; what it asks is only its ontological seat. It says only this: if standard QFT, under the conditions of Lorentz locality, unitarity, and spin-statistics, has already given the CPT theorem [22], then the opposite-arrow 4DD of this framework furnishes precisely a seat compatible with it, rather than issuing a new proof. The ground of that seat lies in the structure of the opposite-arrow 4DD: our side and the opposite side are an involution of exchange-plus-reversal — applied twice, it returns to its start; particle and antiparticle are precisely the two-sided readings of this involution. So a CPT-type symmetry's being robust has, in this framework, its ontological seat, rather than being a symmetry imposed from without.

A clause must be added, lest it be misread as "if robust then all is strictly symmetric": the involution's being exact is so for the whole; while a single branch within it (such as T, the pure reversal of the arrow), when the parameters on the two sides are not strictly equal (after the asymmetry established in the cosmology papers [7]), is no longer strict. Yet why C, P, T may each break, why CP violation appears in the weak interaction, and the precise dictionary between this and the asymmetry of the opposite-arrow 4DD, belong to the four-forces and cosmology series (of one origin with the weak-interaction chirality fixed there by the same asymmetry [8]); this paper retains only a weak commitment — that the involution of the opposite-arrow 4DD is compatible with a CPT-compatible pairing — and computes no quantity of breaking here.

4. The four-forces interface: the grammar by which capacity distributions are glued

This paper interfaces with the four-forces series not by opening with "what kind of thing is a force" — that opening falls straight into the virtual-particle misreading of "what is a mediator." The more proper entry is to ask: how is the field's capacity, between cell and cell, legitimately compared, transported, made coherent, and coupled? This set of rules for "how to glue" is where gauge resides: the connection between adjacent cells is that gluing rule; its curvature is the gluing's mismatch, the discrepancy (holonomy) accumulated by capacity around one closed loop; and what is called the mediator of a force is but an excitable mode of this grammar.

From this, U(1), SU(2), SU(3) and their kind are marked, in this paper, only as the grammar of legitimate transformation of pre-closure capacity, only as the locus of their interface; their respective derivations — whence the gauge group, why three generations, the geometry of the mixing matrix — all belong to the four-forces series [12], cited here and not restated. Gravity must be set apart: it does not stand alongside the other three as one more Yang–Mills force but keeps its station of 4DD read-out and broadcast (after the relativity and information papers of this series); this paper's gauge interface with field theory stops at U(1), SU(2), SU(3), and the field-theoretic treatment of gravity is an effective layer, a boundary case, whose full ontology lies in the relativity and gravity-broadcast papers. What is not treated must also be stated plainly: the Higgs field, electroweak symmetry breaking, Yukawa couplings, the mass spectrum, anomaly cancellation, the three generations and the mixing matrix all belong to the specialist paper on the four-forces / Standard Model interface; this paper treats only the ontological interface at the layer of field capacity and gauge gluing.

Here one may give the example that best displays the framework's strength: the Aharonov–Bohm effect [14]. The electron does not enter the region of the magnetic field, yet acquires a displacement from the phase carried by the vector potential — the standard reading already knows this to be due to the topology of the connection (potential), not the local field strength. In this framework's view, this confirms precisely one sentence: what truly bears the orientation of pre-closure capacity is not merely the local field strength but the gluing grammar itself; the coherent state beneath the slot does, in the form of capacity, occupy the global topology of that multiply-connected space. Written in the grammar of gluing: the phase accumulated around one loop is precisely the global discrepancy (holonomy) of the gluing around a closed loop, Hol_G(γ) = 𝒫 exp ∮_γ G; in the U(1) case, written exp(iq ∮_γ A_μ dx^μ / ℏ) — this paper does not recompute this phase but reads it as the holonomy of the gluing grammar, of one interface with the gauge connection of the four-forces series [12]. More than computing a scattering cross-section, this example displays where the framework is strong.

As for spin and statistics: bosonic commutation and fermionic anticommutation are, in this framework, algebraic facts — they lie across the kinematics of the capacity layer (after the earlier thesis: statistics is the kinematics of the capacity layer, not to be confounded with the vertical axis of pre- and post-closure). Yet that spin-statistics is a theorem [23], its full proof leaning on Lorentz, on microcausality, on positive energy — so this paper establishes here only the layer of "commutation/anticommutation as algebraic fact," and cites the four forces for the full theorem, marking that it rests upon Lorentz; before Lorentz is cashed out, it does not pre-empt spin-statistics.

---

Part III. Classification, Virtual Particles, and Falsifiability

— virtual particles have no ontological name; what can be re-read is the propagator; three "non-contents" must not be conflated

1. Virtual particles: no ontological name; what can be re-read is the propagator

The picture of virtual particles has no ontological standing. Computational convenience (the perturbative expansion, the internal lines of Feynman diagrams) is not ontological actuality; its ground can be stated precisely by the criterion running through what precedes: a real particle is L3-marked-not-constituted — it is on-shell, addressable, closure-ready; whereas a virtual particle, with energy and momentum flying free within the momentum integral, is off-shell, has no addressable real station, no independently recordable detection, and no content born at settlement. Lacking the L3 mark, it has no ontological name.

Yet not to reify the virtual particle is not to judge this stretch as mere "computation." What can be positively re-read is the propagator: it is the correlation kernel that pre-closure field capacity maintains between two boundary queries; the internal line of a Feynman diagram is a term of this correlation, bearing no content. So what should be demoted is the word "virtual particle" and the image of that little ball; while the propagator, Green's functions, and correlation functions are what this paper re-reads. (Two senses must be parted, to accord with the abstract: the propagator as a continuous analytic formal object — the function G itself — belongs to the effective-layer description, of a kind with the formal objects listed in the abstract; what is here re-read as real is the correlation kernel it encodes, namely the correlation of pre-closure capacity between two queries, real-but-effective at tier 2. The two senses must be parted, and the reality of the latter must not lead one to take the former — that continuous analytic form — for the ontology.) Especially in gauge theory, the decomposition of internal lines depends on the gauge choice and the perturbative representation, and non-perturbative formalisms can speak the same physics without borrowing the "virtual particle" image — so the non-ontology of the virtual particle is not merely a philosophical assertion but hews to the practice of field theory. One layer further may be pinned geometrically: the virtual particle lodges in the global integral over a continuous momentum space, whereas this framework's real connectedness is only the local gluing Gᵢⱼ between adjacent cells in position space; to fit that discrete local hopping with a continuous global Fourier transform inevitably overflows, mathematically, into those off-shell terms that are off-shell and nonconserving. So the virtual particle is an artifact not only because it is off-shell but because it parasitizes a "global continuous momentum manifold" that the underlying physics does not possess. The S-matrix likewise — it maps unitarily between a prepared in-state capacity configuration and a detectable out-state event menu; the real out-state content is given by the settlement of the detector record; its internal process is still that pre-closure field capacity, not yet settled (after the foregoing).

2. Three "non-contents" must not be conflated

There are three things in field theory, all of which are "not content," yet each its own and not to be confounded: the vacuum baseline, the virtual-particle internal line, and the remainder of settlement. The vacuum baseline is that ineliminable minimal capacity floor (after the foregoing), and is real; the virtual-particle internal line is a correlation kernel, bearing no content (after the previous section); the remainder of settlement is the unselected term discarded in one closure, headed to the 4DD broadcast conservation ledger (after the two information papers [10][11]).

This distinction answers precisely a question that will naturally be asked: does that remainder of settlement leave a shadow in radiative corrections? The answer is: no. First it must be explained what kind of assertion this "no" is — it is not a new computational result but a discipline of ontological stratification: this paper derives no numerical difference in radiative corrections from the SAE remainder. What radiative corrections probe is the vacuum baseline — the vacuum structure that standard QED long ago computed and that this paper only re-reads; while the remainder of settlement heads to that broadcast channel, not into the radiative structure of field theory. So this paper does not, within field theory, manufacture a separate emission mechanism for the remainder, nor seek its shadow in radiative corrections; the remainder is granted only one destination (into the 4DD broadcast conservation ledger) and is sought nowhere else.

3. Ultraviolet divergence and renormalization

This framework's discrete prior has something to say about ultraviolet divergence. Those ultraviolet infinities in standard field theory arise from integrating the field as continuous down to arbitrarily small scales; and this framework's discrete prior denies precisely this point — so the ultraviolet divergence is, in this framework's view, a formal ailment displayed when the effective-layer continuity is pushed toward arbitrarily short distances, a developing-out of the discrete prior at the field-theoretic layer. Yet this developing-out, while concordant with the discrete prior, does not constitute an independent proof of the discrete prior, nor does it mean that a naive Planck cutoff automatically disposes of renormalization. Renormalization is the process by which the capacity algebra is re-expressed across scales, a matter of coarse-graining.

Yet this developing-out is not "solving": this paper by no means claims "renormalization solved because of discreteness." The effective-field-theory viewpoint of Wilson [15] long ago showed that renormalization is not an ailment that must be "solved" — the cutoff need not be physical, and the running coupling has its own logic of scales. To call "a physical cutoff" "solving the ultraviolet" is both overstatement and a misunderstanding of renormalization. So this paper marks it here only as a framework-layer structural commitment (a developing-out of the discrete prior at the field layer), and stops there; the step of "solving" is held outside the door. Further: to force in a spatial cutoff would invite the Lorentz problem back — whereas this framework's discreteness rests on that real, absolute Planck substrate, with Lorentz carried by the two-layer structure, touching neither horn (after the foregoing).

4. The classification of the remaining pain points

The remaining matters this paper notes, or cites, without unfolding one by one. The indistinguishability of identical particles: in this framework, that is multiple occupation stations on the capacity algebra of one and the same field, not many little things each with a private name — which accords with the main line of "the particle is not fundamental." Spin-statistics: see the foregoing; the conditions of its full theorem lie with Lorentz. Aharonov–Bohm: see the foregoing, as an example of the topology of gluing grammar. Casimir, Lamb shift, anomalous magnetic moment: see above, all as posterior anchors of the vacuum baseline, re-read here, not predicted.

5. The grading of commitments, and how to falsify

This paper finally states plainly the weight of its various claims and where they may be falsified, lest the reader mistake a re-reading for a prediction or an open question for a settled one.

As to weight, this paper grades by strength of commitment into four explicit levels, labeling each, lest the reader mistake an ontological re-reading for an empirical prediction.

First, T1 (inherited empirical–formal commitments): this paper inherits the formalism and observable predictions of standard QFT without alteration; real particles and antiparticles, as detectable on-shell excitations, are inherited from the standard posterior; this paper offers no new measurable predictions in the domain of mature pure field theory.

Second, the a priori level (a framework-level risk commitment, not a T1 empirical result): the fundamental layer is discrete, and the continuous field is an effective layer (after the foregoing and the foundations paper [4]); the specific discrete scale awaits measurement. This is not an empirically established fact but a framework-level, falsifiable a priori risk commitment of SAE.

Third, T2 (ontological articulation, erecting the ontological image without altering observable predictions): the field operator as a capacity query; Fock as an occupation menu; the vacuum as baseline capacity; the gauge connection as the gluing rule; curvature as the gluing's discrepancy around a loop (holonomy); the antiparticle as the read-out, on our side, of the charge-conjugate sector of the opposite-arrow 4DD; entanglement entropy as the cut of boundary gluing capacity (on the vacuum baseline); UV divergence as the developing-out of over-extended continuity; renormalization as the re-expression of the capacity algebra across scales.

Fourth, T3 (programmatic, held outside the door, not feigned as settled): the specific discrete scale; deriving from SAE the gauge group of the Standard Model and the three generations, masses, and mixing; rigorously recovering Lorentz and gauge invariance from discreteness; the precise dictionary between C/P/T and the opposite-arrow 4DD; the origin of CP violation (referred to cosmology and the four forces); the SAE-structural derivation of radiative corrections; and the rigorous construction of Haag's difficulty and the continuum limit.

As to falsification, the claims of this paper may be sorted, by their mode of failure, into three types, listed explicitly below.

Type A (no direct new empirical prediction): this paper offers no new measurable prediction in the domain of mature pure field theory — Schwinger, Casimir, the Lamb shift, the anomalous magnetic moment, and collider scattering all serve as standard posterior anchors, not predictions proprietary to this framework. This paper states forthrightly that it has no nontrivial new prediction in this domain.

Type B (compatibility of structure, may fail on physics): this paper inherits the Lorentz of the relativity series, so its discrete ontology must be compatible with that inherited Lorentz, gauge, locality, and no-superluminal-signaling. Should a strong result show that any fundamental discrete ontology is in principle incompatible with these, then this paper's structure fails; and should the antiparticle reading of the opposite-arrow 4DD prove incompatible with charge conjugation, with CPT, with pair production, and with positive-energy read-out, then this interface fails. What is established here is that discreteness must be compatible with (inherited) Lorentz, not that Lorentz be recovered afresh from discreteness — the latter being a stronger target that this paper does not undertake and that the relativity series likewise handles by inheritance.

Type C (formal re-reading, may fail on itself): should the field operator, Fock, the propagator, and the gauge connection prove unable to be consistently re-read as capacity query, occupation menu, correlation kernel, and gluing rule, without introducing hidden variables, or without reifying the virtual particle as a content-bearer, then this whole formalist re-reading fails; and in particular, should standard phenomena prove inexplicable without reifying the virtual particle as a content-bearer, then that negative characterization fails.

For all three types, this paper lists them forthrightly and welcomes falsification — to push the wall of sighs even a hair's breadth is also a victory for reason.

---

Acknowledgements

This paper, and the SAE framework as a whole, owe much to a long collaboration and crucial discussions with Zesi Chen (陈则思): the a priori commitments of SAE, the universal pattern of the 16DD periodic table, and the overall structure of the framework all profited deeply therefrom.

The writing of this paper was assisted by four AI systems — Claude, ChatGPT, Gemini, and Grok — and by multiple rounds of cross-review; their divergent review comments were of great help in delineating the various boundaries of this paper.

For all that this paper asserts, establishes, or gets wrong, the responsibility rests with the author alone; the contributions of the above collaborators and reviewers end at what is recorded in this section and do not extend to any sharing of responsibility anywhere in the body.

---

References (in order of appearance)

[1] Han Qin. SAE Quantum Mechanics VI: The Born Rule.

[2] Han Qin. SAE Quantum Mechanics VII: Measurement.

[3] Han Qin. SAE Quantum Mechanics VIII: Decoherence and the Emergence of the Classical.

[4] Han Qin. SAE Physical Foundations.

[5] Han Qin. SAE Relativity I: Gravitational Time Dilation (Two-Layer Structure and Planck Absoluteness); SAE Relativity II: The Ontological Unification of Motion Effects and Gravitational Effects.

[6] Han Qin. SAE Information Theory III: The Causal Slot of Information.

[7] Han Qin. SAE Cosmology V: Causal-Law Strengthening, Remainder Conservation, and the Dual-Frame Resolution of the Ġ/G Tension (dual-4DD; Λ₁+Λ₂=0).

[8] Han Qin. SAE Cosmology III: From Λ to a₀ (geometric completion of the dual-4DD asymmetry; weak-interaction chirality fixed by T₁>T₂).

[9] Han Qin. SAE Mass Series: E=Ic³.

[10] Han Qin. SAE Information Theory V: Broadcast.

[11] Han Qin. SAE Information Theory VI: Gravitational-Wave Broadcast.

[12] Han Qin. SAE Four-Forces Series (four-forces–DD mapping).

[13] Stückelberg E. C. G. (1941); Feynman R. P. The theory of positrons. Phys. Rev. 76, 749 (1949).

[14] Aharonov Y., Bohm D. Significance of electromagnetic potentials in the quantum theory. Phys. Rev. 115, 485 (1959).

[15] Wilson K. G., Kogut J. The renormalization group and the ε expansion. Phys. Rep. 12, 75 (1974).

[16] Han Qin. SAE Quantum Mechanics V: Quantum Entanglement (the indecomposable ledger 𝔏).

[17] Han Qin. SAE Information Theory IV: Black-Hole Information via the Causal Spectrum ($R_{\min}(T_H)=2R_s$; Page curve; firewall; black-hole interior). DOI: 10.5281/zenodo.19880112.

[18] Srednicki M. Entropy and area. Phys. Rev. Lett. 71, 666 (1993).

[19] Ryu S., Takayanagi T. Holographic derivation of entanglement entropy from AdS/CFT. Phys. Rev. Lett. 96, 181602 (2006).

[20] Haag R. On quantum field theories. Kgl. Danske Videnskab. Selskab, Mat.-Fys. Medd. 29, no. 12 (1955); Hall D., Wightman A. S. Mat.-Fys. Medd. 31, no. 5 (1957). [Haag's theorem]

[21] Stone M. H. Linear transformations in Hilbert space III. Proc. Natl. Acad. Sci. USA 16, 172 (1930); von Neumann J. Die Eindeutigkeit der Schrödingerschen Operatoren. Math. Ann. 104, 570 (1931). [Stone–von Neumann theorem]

[22] Lüders G. On the equivalence of invariance under time reversal and under particle-antiparticle conjugation for relativistic field theories. Kgl. Danske Videnskab. Selskab, Mat.-Fys. Medd. 28, no. 5 (1954); Pauli W., in Niels Bohr and the Development of Physics, Pergamon (1955). [CPT theorem]

[23] Pauli W. The connection between spin and statistics. Phys. Rev. 58, 716 (1940). [spin–statistics theorem]

[24] Schwinger J. On gauge invariance and vacuum polarization. Phys. Rev. 82, 664 (1951). [pair production in strong fields]

[25] Casimir H. B. G. On the attraction between two perfectly conducting plates. Proc. K. Ned. Akad. Wet. 51, 793 (1948). [Casimir effect]

[26] Lamb W. E., Retherford R. C. Fine structure of the hydrogen atom by a microwave method. Phys. Rev. 72, 241 (1947). [Lamb shift]

[27] Schwinger J. On quantum-electrodynamics and the magnetic moment of the electron. Phys. Rev. 73, 416 (1948). [anomalous magnetic moment, g−2]

第一部分　前闭合场分布的本体

——场不是连续的实体，是离散元胞上的前闭合容量分布；连续的场，只是它在粗粒化层上的一张脸

一、起点

前几篇所论，大抵以一个量子、一次闭合、一道菜单为题。论玻恩规则的一篇[1]给了权重，论测量的一篇[2]给了那一份被读出的单一内容，论退相干的一篇[3]又把退相干之后、落定之前那份失相而仍多容的客观菜单（menu-information）摆到了因果槽之上。可这些，都还是一个对象的前闭合容量——一颗粒子、一具系统、一道叠加。然而自然界并不以单个量子为其常态。当前闭合的容量不再只系于一个对象，而是铺陈在许多元胞、许多模式、许多可被闭合寻址的位置之间时，单粒子的语言便退居为有效的近似，场论的语言由此出现。前篇论退相干，已为由微观转向宏观开了端；本文是它的续篇，所问者：当那份菜单不再只属于一个量子，而属于一整片元胞集合时，它是什么。

本文的辖区须先划清。本文不重做四力系列、不推标准模型、不重述规范结构，只问前闭合的“场分布”在本框架里是何本体、又如何与四力系列的代数与表象相接；至于路径积分、完整的连续极限、经典力学之浮现，皆留与系列收束的一篇。本文待场算符、产生湮灭、真空、虚粒子诸物，是把它们摆回前闭合容量的本体里，不改其任何可观测的预言——本框架在能探测到分野处方出预言，在只复现标准物理处只重读，本篇属后者。

二、离散，与洛伦兹

这条路一迈步，便有一桩须先安顿的事。本框架有一条明摆着的硬先验：任何物理量，在足够小的尺度上，都是离散的（承本系列地基一篇[4]，论退相干的一篇[3]已就此申明，含场、含概率云）。这离散并非一个可有可无的范畴抽象——它有真实的底：本系列论相对论的诸篇[5]早已立下普朗克之绝对——c = l_P/t_P 为普适的贯通率，普朗克之尺度 l_P 作那真实而绝对的自然标度，不变、不随体系而易，此一底层亦不随质量之扭曲而易。须分清的是，这不变而绝对者，是那把尺的刻度（标度 l_P 之为真实）——l_P 乃一普适之常数，其数值如 c 一般对一切观者皆同；这并不等于某一体系里一个元胞的空间广延是 frame-independent。逐元胞真正的洛伦兹不变量，是它所载的因果容量与 4DD 超体积（详见下节）；而元胞之空间广延，下文将见，是随体系而缩的。至于这离散之具体物理标度，是否恰为 l_P、又如何落到场层，乃物理学所当测的问题；本文只把普朗克尺度奉作真实而绝对的自然候选锚点（承论相对论[5]），不替物理学家把它兑现为某一确定之值，其具体离散标度待经验所测（承论退相干[3]、地基[4]）。

而离散素来与洛伦兹不变相忤——一张有其绝对尺度、铺在空间里的网格，似乎自带一个优选的参照系，洛伦兹一变换，网格便露了马脚；可量子场论的洛伦兹不变，已被检验到极高的精度，至今未见破绽。把场读作离散元胞之上之物的框架，于是先须答这一问：那真实而绝对的普朗克离散，如何与洛伦兹不变并存。

此一问，本系列论相对论的诸篇[5]已作交代，本文承之、不复重立。其要在一重两层的结构：底下是普朗克的基底，绝对而普适——光作 2DD 之波、本不载信息，正行于此层，故其速对一切观者皆同；其上是因果槽，随引力、随运动而变其尺度。论相对论的一篇[5]于此立得明白：元胞之空间长度，是随体系而变的——沿运动之向缩为 1/γ；而洛伦兹所不变者，不是那长度，是元胞的 4DD 超体积（R₀³·c·t_P；或于 c=1 之单位下写作 R₀³·t_P）与它所载的每元胞一比特之因果容量。换言之，不变的是容量、是超体积，可变的是长度。这一超体积之不变，论相对论的一篇老实标作标准狭义相对论体积元之不变（雅可比为一）的本体重述——洛伦兹是承之而来、非自先验另推；本文亦承此而不重立。关窍尚有二。其一，那绝对的普朗克基底，是一层本体的基底，不是一个经验上可分辨的优选系：一切惯性的观者，经洛伦兹变换一视同仁，说元胞“在绝对的普朗克尺度上当真缩短”，只是本体的措辞，其所予的可观测效应，仍与标准狭义相对论分毫不差。其二，这基底也不是一张刚硬而有优选方向的方格晶格——若是方格，对角等向便破了洛伦兹的转动与 boost；它毋宁是一张以那守恒之因果容量为节点的关系性的网，正因其所守是容量（超体积）而非某一方向之长度，方能在粗粒化的有效层上兼容洛伦兹之连续对称。故真实而绝对的普朗克离散，与洛伦兹之不变，并行而不相悖。

这里须把两个本不相同的概念分清，免得彼此相累。普朗克之尺度，言那真实而绝对的底层标度；因果槽，则是一道范畴的门槛——一片底层的聚合越过因果闭合之阈、成其为载因果信息者，方称因果槽（承本系列论信息的一篇[6]）。前者绝对而不变，言尺度之真实；后者随条件而变其尺寸，言范畴之跨越。二者各司其义：离散有其真实的普朗克之底，因果槽则是这底之上的一道范畴之分。前篇所立“因果槽乃范畴而非尺寸”，说的正是后者之为范畴——它并不取消前者之为真实；尺寸之非门槛，与尺度之为真实，本是两事。

至于场算符 φ(x) 里那个连续的坐标 x，本文如此安放：它不是基本元胞的私名，而是仪器那一层给出的一张有效的坐标图（chart）。基本的一层，是一片可被闭合寻址的元胞复形（cell-complex），并非预先嵌在某一张固定欧氏空间上的格点；连续的 x，是这片复形在给定仪器与时空背景下被粗粒化所成的图。故 x 属有效的读出层、不属基本的本体层；本文于此不推导诸图之间的协变变换，那属本系列论相对论诸篇所辖。落定在相对论设置下的协变一事——测量不可有跨类空间隔的瞬时坍缩——亦同此辖，本文承其框架而不在此展开。

离散这一头既已立稳——真实而绝对的普朗克之底，因果槽为其上一道范畴之分，洛伦兹之不变由那一重两层的结构所承——下文方才谈得上，这前闭合的场分布，究竟是何本体。

三、承重的本体：离散元胞、局域容量、缝合

既已言明连续的场不是基本的，便须答那基本的一层是什么。这里先须有一把尺，用以分判一物是有效层的描述、还是本体本身——这把尺贯穿全文。在本框架里，分判不系于"数学与物理"之名，而系于"连续与离散"之实：凡连续而莫测之抽象，是有效之层上的描述；凡离散而真实之所在，方是承重之本体——而那承重之离散本体，正落在 3DD（承论相对论[5]之 DD 阶梯：3DD 为质量绑定、空间聚合之所在）。连续的场、算符的代数、传播子与格林函数、路径积分的测度等形式对象，皆连续之抽象，是有效层的描述；离散而真实的元胞、元胞上的容量、元胞间的缝合，才是 3DD 的承重本体。须把话说得公允：算符的代数，是一套表象无关的语法，记着可观测量、记着局域性与对易的结构，自有其极重的分量；本文不贬它，亦不取“连续算符代数即最终本体”那一路——本文只是把那承重者，仍安在离散的元胞容量与缝合之规上，而以代数为其有效的读出形式。连续的代数纵然精当、纵然表象无关，它在本框架里是本体结构之有效语法，而非所立的最终 substrate。

那承重的本体，是这样一片图景：一片离散的元胞——其底即上节所言那真实而绝对的普朗克基底，尚在因果槽之下——元胞各自载着一份局域的、前闭合的 ρ-OR 容量，相邻的元胞之间，有一套缝合的规则——规定局域的容量如何相互比较、如何运输、如何相干、如何耦合。若要给一个精确的骨架：设元胞之集为 C = {cᵢ}，每一元胞 cᵢ 上有一份局域容量代数 𝔄ᵢ 与一份局域 ρ-OR 容量 ρᵢ（此 ρ 非密度矩阵之记号，乃局域容量之记号），相邻元胞之间有缝合规则 Gᵢⱼ；所谓场的分布，不是一组已经实现的内容，而正是这三者 {𝔄ᵢ, ρᵢ, Gᵢⱼ}。那相邻元胞间的缝合，正是日后接到四力时所见的规范联络之所自——此处先按下，论四力的接口一节再表。

连续的场算符 φ(x)，便在这片图景里得了它的位置：它不是“x 那一点上当真弥漫着一团连续的实体”，而是当仪器把这片离散的元胞粗粒化、铺成连续的坐标 x 时，对这份容量分布所发出的一道有效的读出。点 x 本不是基本的元胞，对一点的读出，是粗粒化投影推到极限的产物。所以连续的场，是这份离散容量在有效连续层上的一张脸——它作为有效的构造是真实而有效的，不是基本的底层，亦不只是一种计算的方便。至于那套正则的对易与反对易关系——场论里写作代数的那一套——则是这份容量结构的一种表象无关的描述：它描述的是同一份元胞与容量的本体，而非本体之外另有一物。

顺此可论一桩场论里素称微妙的事——哈格之定理（Haag's theorem）[20]。连续的量子场论，自由度无穷，正则对易关系便有无穷多个幺正不等价的表象，相互作用绘景于是在数学上站不稳——这正是“表象无关”在连续理想化下最吃紧的地方。本框架不声称由 SAE 推导或了结了哈格之定理；哈格所显者，与其说是一桩可被本文直接“解决”的技术小病，不如说是一道结构的警告：勿把那有效连续的相互作用绘景误认作基本的本体。本框架之离散先验于此只作一处有限层的安顿——在有限元胞的正规化、有限的区域、有限个正则自由度之层面，斯通-冯诺依曼型（Stone–von Neumann）[21]的表象唯一性可作一良性之锚（此与任何紫外正规化于有限截断处之所为相通，本文取之作离散先验之一处印证，非自夸独解）；至于连续之极限、无穷体积之极限、与相互作用表象如何严格恢复，仍属系列收束的一篇与技术之专篇。故本文只取它支持一个本体的判断——连续之场非最终之 substrate；而把“严格恢复”那一步，明白地挂在门外。

这一处，也恰好给本文标出一道形式的失败之界（属下文丙型）：倘那有限元胞的容量之网，竟无法给出一个合理的连续归纳极限，或在该极限里不能恢复标准场论之散射、定域、与重整化之可观测，则本文这一套形式主义的重读，便在此失败。哈格之难，于是不是外人攻来的入口，而是本文自陈的、最诚实的一处可证伪。

顺此可向代数量子场论（algebraic QFT, AQFT）递一个明确的接口。AQFT 立一张局域代数之网 O ↦ 𝔄(O)，记着可观测量、定域性与对易之结构；在本框架里，这张网不是“以代数为本体”的稻草人，而恰是那承重的元胞–容量底层在一片有效区域 O 上的可观测之影——可写作 O ↦ CoarseGrain({𝔄ᵢ, ρᵢ, Gᵢⱼ}_{cᵢ⊂O})。故 AQFT 之局域代数，于本框架不被推翻，而被安放进有效之层：它是本体之粗粒化的读出，而非本体本身。此一接口，正可邀治代数场论者循之而入本框架之范畴。

这里把本文的三层本体一次说清，免后文散读：其一，基本而承重者——离散的元胞、元胞上的容量、元胞间的缝合，是 3DD 的真实底层（fundamental）。其二，真实而有效者——连续的场（作有效之结构）、L3 之“粒子”、以及算符之代数（作那承重本体之表象无关的语法、一种有效之描述），皆真而非基（real 而非 fundamental）：它们确是真实之物，却不是最终的底层，而是那底层在有效连续层上的浮现或其语法。其三，纯属表象之伪影者——虚粒子之图像、壳外之内线，是计算之物（artifact），无本体之名（详见第三部分）。须辨一处易混：连续的场（或算符代数），作“有效之结构（语法）”时是真而有效的，作“一个连续的数学对象、或被当作最终 substrate”时则只是有效层之描述、其本身非最终之 substrate——两义须分，不可因前者之真实，而把后者也认作本体。

四、场算符、Fock 与产生湮灭：容量层的读法，及“粒子”之为何物

承上一节，二次量子化那一整套华丽的算符，在本框架里读来便都落回了容量这一层。

场算符所问者，是“若在此处、此模式上设置闭合的寻址，哪些容量可被重组、可被读出”；它是一道查询，不是一团实体。算符取值为分布（operator-valued distribution）的那个“分布”之性，于此也就有了着落：因为点不是基本的元胞，对一点取值，原是粗粒化的仪器投影逼至极限时的说法。

产生与湮灭的算符，于是都不是造物或灭物。a†|0⟩ 在本体上不是从无中生出一颗小球，而是把真空基线里某一模式提升为一份单激发的容量配置——在 1DD 至 3DD 的拓扑特征上作一次局域的重组；唯有当它与探测器、与记录发生那一桩落定时，方才成为单一的内容。湮灭的算符同理：它是这本容量账的重组，不是余项的消灭——余项不可消灭，是本框架一以贯之的律，此处亦然。Fock 的空间，便不是一座排满小粒子的仓库，而是占据数的一张容量菜单 |n₁, n₂, …⟩——可被闭合读出的诸占据数配置之菜单。而那道散射矩阵 S，所演化的正是这些菜单之间的幺正流转，是容量层上的事；落定不在其中——量子场论本身只演化菜单、并不点出单一的内容，点出内容的，是测量、是探测那一桩落定。

这里极易生一桩范畴的错认，须先防住。玻色的多占与费米的独占，看去诱人地像“或”与“与”——玻色之可多占像 ρ-OR，费米之独占（不相容原理）像 ρ-AND——然而这是把两条轴错认作一条。统计这件事，对易与反对易、a†a† 之或为零、不相容之独占，是容量这一层上的运动学：它横在容量层里，规定的是每一模式能容几何。ρ-OR 与 ρ-AND 则是闭合的那一道纵轴：前闭合之多容，与落定之后的单一。纵是费米的场，每一份候选的内容里，某个费米模式仍受不相容原理钉死；ρ-OR 所谓的多容，说的是模式之为多重（modal plurality），并不是说每一模式都可被无限占据。两者不可混为一谈——一混，便要使人误以为本框架的 ρ-OR 天生偏向玻色，那是错的。

那么，既然场算符不造物、Fock 不过是菜单，所谓“粒子”，又是何物？它便不是一个基本的对象。粒子，是一份场的容量花样，在它足以被记录之时，所呈现出的一份合于质壳（on-shell）、随时可被闭合（closure-ready）的真实激发——它已被寻址、它确在场，却尚未构成那一份单一的内容。借本框架的量级阶梯[4]说，粒子是 L3 之“标”而非“构”：是真实的标记（real），不是基本的构成（fundamental），也还不是落定所生的内容。量子场论把粒子整个溶进场算符里、不复以之为基本；本框架不溶解它，而给它一个 L3 标而不构的实位——是真而非基。这一分判看似细，却是后文的枢纽：论虚粒子时，正因其无此寻址、无此实位，方判其无本体之名；论反粒子时，则正凭其是对侧的一份真实激发，而非别物。

五、菜单升到场上：从一物之菜单，到场级的事件菜单

论退相干的一篇[3]所立的那道 menu-information——问之前的可回答性，是容量而非内容——在单系统里，那菜单不过是上与下、自旋之正负、路径之甲乙、位置之某区间与某区间。可一旦升到场上，这菜单便长高了一级：它不再是“某一粒子在何处”，而是整片场构型、或占据构型的一张候选之集。

举对撞之事为例。落定之前，那道场层的菜单，问的不是“某粒子在哪里”，而是种种可能的事件拓扑（event topology）：是两个光子，是一对正负电子，是若干道喷注，还是一道丢失了能量的渠道；退相干、探测器的耦合，使这些事件拓扑成为槽上的菜单——失相，而仍多容；而真正的内容，是某一次探测的记录里，哪一组元胞当真 click 了、能量沉积了几何、径迹又如何连成了一条。

于是场里的 menu-information，便不是“一物之经典菜单”，而是“场级的事件菜单”。这是把退相干一篇的那道菜单，自测量推及于场，而非重述其已言。

六、真空，与零点能

真空，于是也不是空无（nothing）。它是没有任何落定的激发（settled excitation）之时的那一份基线 ρ-OR 容量——全片元胞各处于前闭合的基态，维持着彼此的连通，尚未被任何一处局域的落定所咬合。它甚至不是一个绝对单一的对象：它是相对于某一套闭合寻址之语法而言的“无落定激发”，而非那绝对的、无所谓语法的无——这一层相对之性，到弯曲时空的场论里自会显出分量，此处先按，留与论相对论、论宇宙学的篇章。

零点能，也就不是虚粒子借来还去的戏法。须把不可消去的那一物说准：不可消去的，是真空基线那一份非平凡的容量结构与关联，而不必是某一套模式分解之下零点能的绝对数值本身——那绝对数值，可经正规排序（normal ordering）或重整化而处置，是有效形式主义里的事，并不等于真空基线在本体上为空。零点能，是给定一套模式语法之下，真空基线的一道签名（signature）。此处措辞仍须准：真空是零占据的态，故这一基线是“不可消的最小容量结构”，不是“多容”——多容是别处的事，真空恰恰是那最低的、却仍非零的容量底。

末了须挂一道防火墙。这一层容量的基线，与 4DD 那一层载内容的源——引力之所应者——须分作两层而论。把诸模式的零点能天真地一加、未经层级的审查便径直当作那载内容的源，是不当的；本框架将容量的基线与 4DD 的读出分作两层，正提示这一步不可贸然。故卡西米尔[25]、兰姆移[26]、反常磁矩[27]诸效应，是真空之结构的后验之锚，而非“零点能之绝对数值直接引力化”之据。至于那宇宙学常数究竟几何的问题本身，自有论宇宙学、论引力的篇章去料理，非本文所辖。

七、纠缠之熵，与面积之律：跨界的容量关联

前几节把场的本体落在了那三件 {𝔄ᵢ, ρᵢ, Gᵢⱼ} 上；既如此，纠缠的熵，与那条素称深刻的面积之律，便也在这片图景里得了一个本体的读法，且这读法恰是本框架之长所在。

先须分清两件本不同的事，免与前篇相累。纠缠之为现象——那不可分解的账，A 与 B 合起来的账，不等于各自账之直和（𝔏_AB ≠ 𝔏_A ⊕ 𝔏_B）——论纠缠的一篇[16]已立，本节不重做；本节只问那场论里、区域与区域之间、真空之中的纠缠之熵，是何本体。

其本体，仍在缝合。取一片区域 A 与其外 Ā；A 内的容量与 Ā 内的容量之间的纠缠，是由那些跨过边界 ∂A 的缝合 Gᵢⱼ 担着的——A 里紧贴边界的元胞，与 Ā 里紧贴边界的元胞之间的那一束容量关联。约化到 A、求一道迹、算它的熵（习惯写作 S_A，其约化算符记作 ρ_A；此 ρ_A 是标准的约化密度矩阵，与前文那局域容量之 ρᵢ、与 ρ-OR，皆非一物，勿混），所量的，正是这一束被边界切断的缝合。（又须一言以承前节：§二既言元胞之空间广延随系而缩，则此处“边界元胞之数”与“面积”亦一同随系而变换，所量之 S_A 是协变之量；面积之律并不暗设元胞广延为 frame-independent。）

由此，那条面积之律——纠缠熵之领头项，正比于边界之面积，而非区域之体积——在本框架里得一几何的读法（此读法本文标为 T2 之本体重读，非一普遍之定理）。其理须连两个前提一同说明：其一，缝合是局域的，只在相邻的元胞之间；其二，所论者是真空之基线（或低激发之态）。于此二前提之下，A 与 Ā 之间主要的纠缠之通道，便是那贴着边界的一层元胞，而边界元胞之数，正比于 ∂A 之面积、不正比于 A 之体积。须明言：单凭“局域缝合”一项，于图论上只逼得一个纠缠之上界（min-cut 之限）；若处一高激发或热态，纵在局域缝合之网上，其纠缠之熵亦可服从体积之律。故面积之律之所以显现，是局域缝合配上真空基线之低激发拓扑，而非局域缝合一项独力所逼。可量化地标出这上界：以 χᵢⱼ 记跨界缝合 Gᵢⱼ 之有效容量维数，则由局域跨界缝合所担的领头纠缠，有 S_A ≤ Σ₍⟨i,j⟩∈∂A₎ log χᵢⱼ；若边界之缝合大抵同质（χᵢⱼ≈χ）且近饱和，则领头项 ~ |∂A|·log χ，正比于边界元胞之数。标准场论里费一番手脚方得的 S ~ 面积/ε^{d−2}（自由场之纠缠熵，斯雷德尼基之所算[18]），在本框架里是缝合之局域性（于真空基线上）的一个直接的影。又须界定其所不及：临界系统（如 1+1 维之对数标度）、费米面所致之面积律破坏、规范场之边缘模、拓扑纠缠熵、长程相互作用诸般修正，皆系乎具体之场论与多体之条件，非本节所断——本节只立真空、局域、短程之领头面积项这一 T2 之读。

而那 S_A 里的紫外发散——ε → 0 时的那个无穷——又是一处显影，与论紫外的一节（第三部分）同源：它正是连续之极限下“边界元胞之数 → ∞”；离散的元胞，给了它一个天然的底。纠缠之熵，恰是场论里最干净的一个紫外发散量；本框架于此只标其显影、不算其系数（亦不论无质量场在低维所带的对数修正），纪律与论重整化一节同。

这一节，也把真空那一节落到了实处。§六说真空“维持着彼此的连通”；那连通量到一片区域的边界上，便正是真空的纠缠之熵。真空不是空无、其基线的容量结构是真实的——这一句，于纠缠之熵处得了一个具体的形：真空的纠缠熵，量的是那跨越边界、维系着两侧连通的缝合。

末了须挂一道防火墙。纠缠熵正比于面积——黑洞之熵 S = A/4 亦正比于面积——这一处共振诱人。本框架里，纠缠熵之面积律来自缝合之局域性（于真空基线上）；至于黑洞视界这一特殊之边界所产生的熵、Page 曲线、island 之结构，及其与标准 replica wormhole 计算之异同，则属论黑洞信息的一篇[17]与论相对论诸篇之所辖——全息纠缠熵之标准基准（Ryu–Takayanagi[19]）亦在此被明划于本文之外，本节不入其辖。本文只处理平直空间与一般区域之纠缠之熵与面积之律，至此为止：不推 S = A/4、不算 Page 曲线、不碰 island 与 replica wormhole。那由黑洞细致纠缠所引出的、与标准 island 图景之异同，是一桩留待将来（若有论量子信息的系列与论黑洞信息的一篇相对接）去厘清的开放问题，本文只立此界、不入其中。

---

第二部分　相对论接口与四力接口

——反粒子是对侧的真实激发，不是落定所舍的余项；四力是容量分布相缝合的语法

一、反粒子：对侧 4DD 的真实激发

量子场论较前几篇所论真正多出来的，是相对论。相对论一进来，便带出了反粒子——正电子、反质子之属，皆有其后验的实在，可探测、可记录。本框架于此有一个现成的本体出口，且这出口并非为场论临时所造，而是本系列论宇宙学的一篇[7]早已立下的：3DD 之对称，生出一对时间箭头相反的 4DD；我们这一侧，与那对侧，参数互换而箭头相背。反粒子，便住在那对侧——它是对侧 4DD（相反箭头）那一共轭扇区里的一份真实的、合于质壳的激发，经本侧的仪器读出时，呈为电荷共轭、正能的一份内容。

然而反粒子不是“同一颗粒子在我们这侧的时间里当真倒着跑”。可以借费曼与施图克尔贝格[13]那个“反粒子如同逆时传播的粒子”的老读法作一类比——它确是一个有用的形——然本框架的本体不在“本侧时间里倒放的一段电影”，而在那对侧 4DD 的共轭扇区于本侧读出中的现身。成对产生（pair production）于是也不是从无中凭空造出两个实体：在强外场或背景的耦合之下，真空那一层基线里一对共轭的容量扇区，被提升为可被闭合寻址的激发，本侧读作一个粒子加一个反粒子；其守恒——电荷、能量、动量——由表象层的约束担着。施温格（Schwinger）的强场成对[24]适作一例，然它是标准的现象，本文重读、不另出预言。

这里有一桩值得正面说出的契合，然须把它的分量标准。那对侧 4DD 的结构，当初是为论宇宙学的时间箭头、为那一系列的先验而立的[7]，并不为粒子物理而设。本文不以它替代标准量子场论对反粒子的推导（那须一套完整的相对论场表示）；它所给的，是一个本体的位点——凡标准量子场论已给出的电荷共轭扇区，在本框架里自然落到那对侧 4DD 的共轭扇区上。于是正电子之实有（一九三二年便已坐实的后验），在本文里不是被重新预言，而是与一个独立建立的宇宙学结构发生了契合（consilience）。一个为宇宙学箭头而立的先验结构，与一个粒子物理的后验事实，在此相遇而相印——这不是一桩新的预言，而是一桩先验与后验之间的非平凡契合，本文如实标出而不夸之为预言。又须补一道界：此处所谓“反粒子”，严格指电荷共轭之表象扇区；于自共轭之表示（如光子、π⁰，乃至设若存在的马约拉纳中微子），其共轭扇区可与本侧之读出重合，并不另要求一个经验上判然有别的第二粒子。

由这结构，还落得一条可证伪的咬合，然其向须标准：那对侧 4DD 之读法，要求凡可被标准量子场论确认为粒子–反粒子之共轭对的扇区，皆须能在本框架里落为对侧之共轭配对；倘此读法竟容纳不下电荷之共轭、质量之相等、正能之读出与成对之产生，则此一接口失败。所压者，是那对侧 4DD 接口之能否承接（承论宇宙学的一篇[7]，非本文新造，亦非本文独立预言质量之相等）。可证伪的咬合点，正在此处。

这一相认，须与三事划清界限。其一，CPT 之为定理、粒子与反粒子质量何以严格相等之推导，属四力与相对论之专责，本文不推导——本文所给者是其本体之由（为什么 CPT 合理），非其推导。其二，物质与反物质何以不对称、重子何以生成、我们这一侧何以由物质所主，属论宇宙学之系列（那对侧 4DD 之所自），本文不及。其三，反粒子不是落定的余项，下节专说。

二、反粒子不是落定的余项

论退相干的一篇[3]立过一桩：落定当场点出单一的内容，而那些未被取中的诸项并不消灭，作余项入 4DD 的广播守恒账。读者至此极易生一个危险的误认：既然反粒子住在那对侧、落定所舍的也是那受相反箭头影响的一侧，二者岂非同一出口？

不是。二者同在那对侧 4DD 的几何位点上，却是位点上两个全然不同的物。反粒子，是那对侧扇区里一份真实的、可被本侧探测的激发——它是一桩运动学的实有，可写入 Fock、可成菜单、可被落定读出。落定的余项，则是一次闭合事件里那些未被实现的容量之去向——它是一桩记账的剩余，循论信息的两篇[10][11]所立的渠道，去往 4DD 的广播守恒账（其能量之一面，经质量系列 E=Ic³ 之所立[9]）。一个是扇区之本体，一个是事件之记账剩余；一个可探测，一个只在那全局的守恒账上记着、不发可分辨的引力波。

若把这两者混作一谈，后果是两种灾难性的误读：其一，以为每一次测量里未被取中的那一支，都化作了反粒子——那么做一回斯特恩-格拉赫的实验，便要喷出一堆正电子来，荒谬之至；其二，反过来把反粒子贬作“未实现之可能”的残渣，而不认其为可探测的真实之物——那是对正电子那一类后验事实的不敬。故二者须分明：反粒子是对侧的真实激发（运动学），落定余项是单向化所挤出的记账剩余（动力学、去广播）；同在对侧之几何，是不同之物。

三、CPT 的本体座位，而非定理之推导

本文不推 CPT 之定理，那是物理学家的活；本文所问的，只是它的本体座位。本文只说这一句：若标准量子场论已在洛伦兹之定域、幺正、自旋统计诸条件下给出了 CPT 之定理[22]，则本框架的那对侧 4DD，恰为它提供一个与之相容的本体座位，而非另出一条新证。其座位之由，在那对侧 4DD 的结构：我们这侧与那对侧，是一桩互换加反向的对合——施之再施，复归其初；粒子与反粒子，正是这一对合的两侧之读。故一个 CPT 型的对称之为稳健，在本框架里有其本体的座位，而不是一条外加的对称。

须补一句，以免误读为“既稳健则一切皆严格对称”：那对合之为精确，是就其整体而言，而其中单独的某一支（如纯箭头之反的 T），在两侧参数本不严格相等之时（承论宇宙学诸篇所立的那一点不对称[7]），便不复严格。然 C、P、T 各自何以可破、CP 破缺何以现于弱相互作用、以及它与那对侧 4DD 不对称之精确字典，皆属四力与论宇宙学之系列（与那里已立的、由同一不对称所定的弱力手征一事[8]同源）；本文只留一个弱的承诺——那对侧 4DD 的对合，与一个 CPT-相容的配对相容——而不在此计算任何破缺之量。

四、四力的接口：容量分布相缝合的语法

本文与四力系列相接，不从“力是什么东西”开局——那一开局便要落入“传递子是何物”的虚粒子误读。更妥的入处，是问：场的容量，在元胞与元胞之间，如何被合法地比较、运输、相干、耦合。这一套“如何缝合”的规矩，便是规范之所在：相邻元胞间的联络（connection），即那缝合之规则；其曲率（curvature），即缝合之不相符、容量沿闭合回路一周所积的落差（holonomy）；而所谓力之传递子，不过是这套语法之可被激发的模式。

由此，U(1)、SU(2)、SU(3) 之属，在本文里只标其为前闭合容量之合法变换的语法、只标其接口之所在；它们各自的推导——规范群从何而来、何以三代、混合矩阵几何——皆属四力系列[12]，本文一概 cite 而不重述。引力则须单列：它不与那三者并列为又一个杨-米尔斯之力，而仍守其 4DD 读出与广播之位（承本系列论相对论、论信息诸篇）；本文与场论之规范接口，到 U(1)、SU(2)、SU(3) 为止，引力的场论处置是有效层、是边界情形，其完整本体在论相对论与论引力广播的篇章。又须明言所不处理者：希格斯之场、电弱之对称破缺、汤川之耦合、质量之谱、反常之相消、三代与混合矩阵，皆属四力与标准模型接口之专篇；本文只处理场之容量与规范之缝合这一层的本体接口。

此处可举一个最见本框架之长的例子：阿哈罗诺夫-玻姆之效应[14]。电子并未进入磁场之所在，却因那矢量势所带的相位而生位移——标准的读法已知其为联络（势）之拓扑、而非局域之场强所致。在本框架看来，这恰好坐实了一句：真正承载前闭合容量之取向的，不只是局域的场强，而是那缝合的语法本身；槽下的相干之态，确以容量的形式，占着那多连通空间的整体之拓扑。以缝合之语法写之：那绕行一周所积之相位，正是缝合沿闭合回路之全局落差（holonomy），Hol_G(γ) = 𝒫 exp ∮_γ G；于 U(1) 之情形，便写作 exp(iq ∮_γ A_μ dx^μ / ℏ)——本文不重算此相，只把它读作缝合语法之 holonomy，与四力系列之规范联络[12]同属一个接口。比起算一个散射的截面，这个例子更见本框架之所长。

至于自旋与统计：玻色之对易、费米之反对易，在本框架里是代数的事实——它们横在容量这一层的运动学里（承前文所立：统计是容量层之运动学，与前后闭合那道纵轴不可混）。然自旋统计之为定理[23]，其完整之证，须倚洛伦兹、倚微因果、倚正能——故本文于此只立“对易/反对易乃代数事实”这一层，完整的定理则 cite 四力、并明其压在洛伦兹之上；洛伦兹未兑现之前，不抢自旋统计。

---

第三部分　归类、虚粒子，与可证伪

——虚粒子无本体之名，可重读者是传播子；三种“非内容”不可混同

一、虚粒子：无本体之名，可重读者是传播子

虚粒子之图像，无本体论之地位。计算之便（微扰之展开、费曼图之内线）不等于本体之实有；其由可循前文那条贯穿的判准说准：实粒子是 L3 之标而不构——它合于质壳、可被寻址、随时可闭合；而虚粒子在动量积分里能量动量乱飞，本不合于质壳（off-shell），无那一份可被寻址的实位，亦无可独立记录的探测、无落定所生的内容。无 L3 之标，故无本体之名。

然不坐实虚粒子，并非把这一截只判作“计算”而已。可正面重读的，是传播子（propagator）：它是前闭合的场容量，在两次边界查询之间所保持的那一段关联之核（kernel）；费曼图的内线，是这关联之项，不载内容。故该被降格的，是“虚粒子”这个词与那个小球的图像；而传播子、格林函数、关联函数，才是本文重读之所对。（须辨两义，以合摘要：传播子作一个连续解析的形式对象——那函数 G 本身——属有效层之描述，与摘要所列之形式对象同；此处所重读为真者，是它所编码的那一段关联之核，即前闭合容量在两次查询之间的关联，乃真而有效之 tier 2。两义须分，不可因后者之真，而把前者那连续解析之形式也认作本体。）尤在规范场论里，内线之分解依乎规范之选取与微扰之表示，非微扰之形式亦可不借“虚粒子”之图像而说同一物理——故虚粒子之无本体，不止是一句哲学的论断，更贴着场论之实操。更可从几何上再钉一层：虚粒子寄身于连续动量空间之全域积分，而本框架之真实连通，只在位置空间相邻元胞间之局域缝合 Gᵢⱼ；以连续之全局傅里叶变换去拟合那离散之局域跳变，数学上便不免溢出那些不合质壳、不守恒的壳外之项。故虚粒子之为伪影，不止因其 off-shell，更因其寄生于一个底层物理并不拥有的“全局连续动量流形”。S 矩阵亦然——它在制备好的入态容量配置与可探测的出态事件菜单之间作幺正之映，真正的出态内容，由探测器记录之落定给出；其内部之过程，仍是那前闭合的场容量，未曾落定（承前文）。

有三事须各别、不可相混。其一，虚粒子不是粒子：它 off-shell、无独立的探测记录、无落定的内容。其二，虚粒子不是余项：它不是某一次闭合所舍的未取之项。其三，虚粒子不是真空之涨落：卡西米尔、兰姆移、反常磁矩背后那真实的真空结构之效应，是真实的，不是微扰记账的伪影；勿因“虚粒子乃伪影”一句而误伤了真空之实有。

此一断言，留与经验去验：倘有朝一日，某一标准的现象，非把虚粒子坐实为一个载内容之物便不能解释，则它即告失败。本文不以为然，然欢迎此种证伪。

二、三种“非内容”不可混同

场论里有三样东西，都“不是内容”，却各是各、不可混作一谈：真空之基线、虚粒子之内线、落定之余项。真空之基线，是那不可消的最小容量之底（承前文），是真实的；虚粒子之内线，是关联之核、不载内容（承上节）；落定之余项，是一次闭合所舍的未取之项，去往 4DD 的广播守恒账（承论信息两篇[10][11]）。

这一分清，恰好回答一个自然会被问起的问题：那落定的余项，可会在辐射修正里留下影子？答曰：不会。须先说明这“不会”是何种断言——它不是一个新的计算结果，而是本体分层的纪律：本文不从 SAE 的余项推出任何辐射修正的数值差异。辐射修正所探的，是那真空之基线——它是标准量子电动力学早已算得、本文只作重读的真空结构；而落定之余项，去的是那广播之信道，不进场论之辐射结构。故本文于场论之内，不另造一套余项之发射机制，亦不去辐射修正里寻它的影子；余项只认其一条去向（入 4DD 广播守恒账），别处不寻。

三、紫外之发散与重整化

本框架之离散先验，于紫外之发散一事，有一处可说。标准场论里那些紫外的无穷大，源于把场当作连续到任意小的尺度去积分；而本框架之离散先验，正不认这一点——故那紫外之发散，在本框架看来，是有效层之连续被推向任意短的距离时所显出的一处形式的病灶，可作离散先验在场论层之一处显影。然这显影与离散先验相合，却不构成离散先验的一个独立证明，亦不意味着一个素朴的普朗克截断便自动了结了重整化。重整化，则是容量之代数随尺度而重新表述之过程，是粗粒化之事。

然这一处显影，并不等于“解决”：本文绝不声称“因离散故解决了重整化”。威尔逊[15]之有效场论的视角早已表明，重整化本不是一桩须被“解决”之病——截断未必是物理的，跑动的耦合自有其层级之理。把“物理之截断”说成“解决了紫外”，既是越言，亦误解了重整化。故本文于此只标其为一桩框架层之结构承诺（离散先验在场层之显影），到此为止；那“解决”的一步，挡在门外。又：硬塞一个空间的截断，反会把洛伦兹之问题请回来——而本框架之离散，是落在那真实而绝对的普朗克之底、由两层结构承着洛伦兹的，两头皆不沾（承前文）。

四、其余诸痛点之归类

余下数事，本文点到、或 cite，不一一展开。同种粒子之不可分辨：在本框架里，那是同一片场之容量代数上的多个占据之位，不是多个各有私名的小物——这正合“粒子非基本”之主线。自旋统计：见前文，其完整定理条件于洛伦兹。阿哈罗诺夫-玻姆：见前文，作缝合语法之拓扑一例。卡西米尔、兰姆移、反常磁矩：见上文，皆作真空基线之后验锚，本文重读、非预言。

五、承诺之分级，与如何证伪

本文最后把诸般主张的分量、与其可证伪之处，一并交代清楚，免读者把重读误作预言、把开放误作已解。

就分量而言，本文按承诺之强度明分四级，显标其级，免读者把本体之重读误作经验之预言。

其一，T1（继承之经验–形式承诺）：本文继承标准量子场论之形式与其可观测之预言、不加更改；实粒子与反粒子之为可探测的质壳激发，乃继承标准之后验；本文于成熟纯场论之域不出新的可测之预言。

其二，先验级（框架之风险承诺，非 T1 之经验结果）：基本之层为离散、连续之场为有效之层（承前文与地基[4]）；其具体之离散标度待测。此非一经验已证之事，而是 SAE 一项框架级、可证伪的先验风险承诺。

其三，T2（本体之 articulation，立本体之像而不更可观测之预言）：场算符之为容量之查询、Fock 之为占据之菜单、真空之为基线容量、规范联络之为缝合之规、曲率之为缝合沿回路之落差（holonomy）、反粒子之为对侧 4DD 共轭扇区于本侧之读出、纠缠熵之为边界缝合容量之切断（于真空基线上）、紫外发散之为连续被过度延展之显影、重整化之为容量代数随尺度之重述。

其四，T3（纲领，挂在门外、不佯称已解）：具体之离散尺度、自 SAE 推出标准模型之规范群与三代质量混合、自离散严格地恢复洛伦兹与规范之不变、C/P/T 与对侧 4DD 之精确字典、CP 破缺之起源（归论宇宙学与四力）、辐射修正之 SAE 结构推导、哈格之难与连续极限之严格构造。

就证伪而言，本文之主张可按其失败的方式分作三型，明列于下。

甲型（无直接新经验预言）：本文于成熟的纯场论之域，不出新的可测之预言——施温格、卡西米尔、兰姆移、反常磁矩、对撞之散射，皆作标准之后验锚，不作本框架独有之预言。本文据实声明其于此域无非平凡之新预言。

乙型（结构之相容，可败于物理）：本文继承相对论系列之洛伦兹，故其离散之本体，须与那继承而来的洛伦兹、规范、定域、与不可超光速之传讯相容。倘有强的结果表明，任何基本离散之本体原则上皆无法与这些相容，则本文之结构失败；又，倘那对侧 4DD 之反粒子读法，竟与电荷共轭、与 CPT、与成对产生、与正能之读出不能相容，则此一接口失败。此处所立，是离散须与（继承的）洛伦兹相容，而非从离散去重新恢复洛伦兹——后者是一个本文并不承担、相对论系列亦以继承处置的更强之靶。

丙型（形式之重读，可败于自身）：倘场算符、Fock、传播子、规范联络，竟不能被一致地重读为容量之查询、占据之菜单、关联之核、缝合之规，而非引入隐变量、或非把虚粒子坐实为载内容之物不可，则本文这一套形式主义的重读失败；尤其，倘非把虚粒子坐实为载内容之物便不能解释标准之现象，则本文那一负面之定性失败。

凡此三型，本文皆如实列出，欢迎证伪——推那叹息之墙挪动一毫一厘，也是理性的胜利。

---

致谢

本文及 SAE 框架之整体，得力于与陈则思（Zesi Chen）长期之合作与关键之讨论：SAE 之先验承诺、16DD 周期表之普遍模式、框架之总体结构，皆深得益于此。

本文之写作，借助 Claude、ChatGPT、Gemini、Grok 四个 AI 系统之协助与多轮交叉评审；其分歧的评审意见，于本文诸般界限之厘定，助益良多。

凡本文之所断、所立、所误，文责尽归作者一人；上述协作者与评审者之贡献，止于此节所记，不延及正文任何处之责任分担。

---

参考文献（按出现序）

[1] 秦汉. SAE 量子力学 VI：玻恩规则.

[2] 秦汉. SAE 量子力学 VII：测量.

[3] 秦汉. SAE 量子力学 VIII：退相干与经典之涌现.

[4] 秦汉. SAE 物理学基础（地基篇）.

[5] 秦汉. SAE 相对论 I：引力时间膨胀（两层结构与普朗克绝对）; SAE 相对论 II：运动效应与引力效应的本体论统一.

[6] 秦汉. SAE 信息论 III：信息的因果槽.

[7] 秦汉. SAE 宇宙学 V：因果律加强、余项守恒与 Ġ/G 张力的双 frame 消解（双 4DD、Λ₁+Λ₂=0）.

[8] 秦汉. SAE 宇宙学 III：从 Λ 到 a₀（双 4DD 不对称的几何完备；弱力手征由 T₁>T₂ 定）.

[9] 秦汉. SAE 质量系列：E=Ic³.

[10] 秦汉. SAE 信息论 V：广播.

[11] 秦汉. SAE 信息论 VI：引力波广播.

[12] 秦汉. SAE 四力系列（四力–DD 映射）.

[13] Stückelberg E. C. G. (1941); Feynman R. P. The theory of positrons. Phys. Rev. 76, 749 (1949).

[14] Aharonov Y., Bohm D. Significance of electromagnetic potentials in the quantum theory. Phys. Rev. 115, 485 (1959).

[15] Wilson K. G., Kogut J. The renormalization group and the ε expansion. Phys. Rep. 12, 75 (1974).

[16] 秦汉. SAE 量子力学 V：量子纠缠（不可分解之账 𝔏）.

[17] 秦汉. SAE 信息论 IV：黑洞信息通过因果谱（$R_{\min}(T_H)=2R_s$；Page 曲线；firewall；黑洞内部）. DOI: 10.5281/zenodo.19880112.

[18] Srednicki M. Entropy and area. Phys. Rev. Lett. 71, 666 (1993).

[19] Ryu S., Takayanagi T. Holographic derivation of entanglement entropy from AdS/CFT. Phys. Rev. Lett. 96, 181602 (2006).

[20] Haag R. On quantum field theories. Kgl. Danske Videnskab. Selskab, Mat.-Fys. Medd. 29, no. 12 (1955); Hall D., Wightman A. S. Mat.-Fys. Medd. 31, no. 5 (1957). [Haag's theorem]

[21] Stone M. H. Linear transformations in Hilbert space III. Proc. Natl. Acad. Sci. USA 16, 172 (1930); von Neumann J. Die Eindeutigkeit der Schrödingerschen Operatoren. Math. Ann. 104, 570 (1931). [Stone–von Neumann theorem]

[22] Lüders G. On the equivalence of invariance under time reversal and under particle-antiparticle conjugation for relativistic field theories. Kgl. Danske Videnskab. Selskab, Mat.-Fys. Medd. 28, no. 5 (1954); Pauli W., in Niels Bohr and the Development of Physics, Pergamon (1955). [CPT theorem]

[23] Pauli W. The connection between spin and statistics. Phys. Rev. 58, 716 (1940). [spin–statistics theorem]

[24] Schwinger J. On gauge invariance and vacuum polarization. Phys. Rev. 82, 664 (1951). [pair production in strong fields]

[25] Casimir H. B. G. On the attraction between two perfectly conducting plates. Proc. K. Ned. Akad. Wet. 51, 793 (1948). [Casimir effect]

[26] Lamb W. E., Retherford R. C. Fine structure of the hydrogen atom by a microwave method. Phys. Rev. 72, 241 (1947). [Lamb shift]

[27] Schwinger J. On quantum-electrodynamics and the magnetic moment of the electron. Phys. Rev. 73, 416 (1948). [anomalous magnetic moment, g−2]