Programme note · 16 November 2026 · Nanjie Ma ·
ORCID 0009-0002-4415-1209
About / Work subpage. Not a homepage hero. Not a ranking. Not a breakthrough notice. Not cosmogenesis payment. Updates the 15 September 2026 on-record disclosure. Companion 26 is on record as of 23 September 2026. Companions 27 and 28 are on record as of 13 October 2026. Companion 29 is on record as of 14 October 2026. Companions 30 and 31 are on record as of 15 October 2026. Companion 32 is on record as of 18 October 2026. Companions 33 and 34 are on record as of 20 October 2026. Companions 35 and 36 are on record as of 23 October 2026. Companion 37 is on record as of 23 October 2026; version 2 is on record as of 14 November 2026. Companion 38 is on record as of 27 October 2026. Companion 39 is on record as of 28 October 2026. Companions 40 and 42 are on record as of 2 November 2026. Companions 47, 48 and 49 are on record as of 10 November 2026. Companion 50 is on record as of 11 November 2026; version 2 is on record as of 14 November 2026. Companion 51 is on record as of 15 November 2026; version 2 is on record as of 15 November 2026; version 3 is on record as of 16 November 2026. Companion 52 is on record as of 16 November 2026; version 2 is on record as of 16 November 2026; version 3 is on record as of 16 November 2026. Companion 53 is on record as of 16 November 2026. Companion 54 is on record as of 16 November 2026. Companion 55 is on record as of 16 November 2026. Companion 56 is on record as of 16 November 2026. Companion 57 is on record as of 16 November 2026. Companion 24’s DOI is listed in the deposit table. Companion 41 is not listed.
Covering kernel Spectral algebra Negatives Form note Independent
What is pinned is discrete generally covariant coherent dynamics on this kernel. The object is in place, the classification is in place, and which principles pin what — and which do not — are written down. Foundational work asks from whom this family of questions first became a mathematical object. It does not require that the universe already be derived, and it does not require that every later layer already be paid.
The counter-intuitive sentence: general covariance plus coherence does not pin the birth share. The share is boundary data, not something covariance derives.
In late August I wrote the minimal covering kernel. There are only two rules: a birth has either one parent, or exactly two incomparable parents. In hand there was a four-event diamond, enough to show it is not a forest. Consecutive work through September then showed what this kernel can actually pin.
At a finite horizon the covariant coherent dynamics is a simplex; the extreme points are the vertices. At each vertex the number of two-parent births is determined, with zero variance. At layer 12 there are 349 391 such extremal covariant dynamics. So it is not that one good principle is missing: the whole covariance class cannot, in principle, select that share.
The infinite side is harder. Even tail-triviality and ergodicity do not pin it: every density in the closed interval from 0 to 1 is realised at an extreme point, and each density still has infinitely many pure phases behind it. The certified trajectories are all ladders; the longer they are, the closer they look to a total order. The unique fixed point of the coarse-graining semigroup is the chain. The chain is not the universe. It is the ruler this kernel itself emits.
Parent paper · Zenodo Companion 20 · proofs begin here Companion 25 · Zenodo Companion 26 · Zenodo Companion 27 · Zenodo Companion 28 · Zenodo Companion 29 · Zenodo Companion 30 · Zenodo Companion 31 · Zenodo Companion 32 · Zenodo Companion 33 · Zenodo Companion 34 · Zenodo Companion 35 · Zenodo Companion 36 · Zenodo Companion 37 · Zenodo Companion 38 · Zenodo Companion 39 · Zenodo Companion 40 · Zenodo Companion 42 · Zenodo Companion 47 · Zenodo Companion 48 · Zenodo Companion 49 · Zenodo Companion 50 · Zenodo Companion 51 · Zenodo Companion 52 · Zenodo Companion 53 · Zenodo Companion 54 · Zenodo Companion 55 · Zenodo Companion 56 · Zenodo Companion 57 · Zenodo ORCID
| Record | Role | Deposit |
|---|---|---|
| Parent manuscript | Programme text | 10.5281/zenodo.23729585 |
| Companion 20 | Finite-horizon covariant classification; proofs begin here | 10.5281/zenodo.22524084 |
| Companions 21–22 | Ergodic arity spectrum; certified ladder locus | 10.5281/zenodo.22520894 |
| Companion 23 | Measure-level thinning semigroup on the locus | 10.5281/zenodo.22546067 |
| Companion 24 | Labeled-path birth-prefix coarse map; different object, not covering-kernel classification | 10.5281/zenodo.22648199 |
| Companion 25 | Geometry of the covering mixed-extremality bar; not payment | 10.5281/zenodo.22769118 |
| Companion 26 | Two-chain last closed form; liminf inhabitation; not leftover r | 10.5281/zenodo.22919449 |
| Companion 27 | Order polytopes, port scaling, finite non-identifiability; not continuum geometry | 10.5281/zenodo.23334925 |
| Companion 28 | Ordering-fraction normalisation; covering birth is not sequential growth; not a later interpolating constructor | 10.5281/zenodo.23335702 |
| Companion 29 | Longest-chain height and Dilworth width of covering orders versus sequential measures; not a uniqueness growth-law | 10.5281/zenodo.23355969 |
| Companion 30 | Covering-seed manifold-likeness exclusion; sprinkling is a comparator, not a programme affiliation; not a uniqueness theorem | 10.5281/zenodo.23385450 |
| Companion 31 | Incomparability ortholattice of covering histories is not quantum logic; standard functor; not spacetime | 10.5281/zenodo.23386149 |
| Companion 32 | History-independent Hasse-degree recognition of covering-process orders; not uniqueness of coupled reconstruction at all sizes | 10.5281/zenodo.23448535 |
| Companion 33 | Sequential extra-input interpolants of covering orders versus two-dimensional sprinkling; not a generation rule | 10.5281/zenodo.23489387 |
| Companion 34 | Handshake surplus of conservative extra-input on covering-process stems; leftover ports are vertex-local; not later laws | 10.5281/zenodo.23489477 |
| Companion 35 | Convex quotients of covering-process orders; leftover Hasse degrees do not increase; not a generation rule | 10.5281/zenodo.23596073 |
| Companion 36 | Covering-process evolution is to a family; interpolants need not force an instance | 10.5281/zenodo.23596461 |
| Companion 37 | Canonical path-entropy measure; all-cardinality T1 envelope; complete size-10 reconstruction; not coupled reconstruction for all sizes | 10.5281/zenodo.24219247 |
| Companion 38 | Exponential path-entropy family; sector-balanced local growth; Dilworth-Turan coarse quotients; symplectic Berry holonomy; not spacetime | 10.5281/zenodo.23685033 |
| Companion 39 | Poset limits of three named covering growth laws; not spacetime; not Lorentz invariance | 10.5281/zenodo.23731189 |
| Companion 40 | Young sector of covering growth; Plancherel descent of Young-diagram covering histories; not spacetime; not Lorentz invariance | 10.5281/zenodo.23879513 |
| Companion 42 | Greene–Kleitman shadow and two-spine lift of covering growth; labeled 1+1 sprinkling representation; not spacetime; not a remint of Companion 36 | 10.5281/zenodo.23879838 |
| Companion 47 | Full-support central measure on the covering growth graph; seed-rigid-ray mixtures fail; not spacetime | 10.5281/zenodo.24099142 |
| Companion 48 | Unique-token extras and ultrahomogeneous embedding squeeze; not leftover interpolating sequential type; not spacetime | 10.5281/zenodo.24099493 |
| Companion 49 | Covering-native operator geography of the covering-law family; not leftover interpolating sequential type; not spacetime | 10.5281/zenodo.24099767 |
| Companion 50 | Left identity and complementary negative directions of the covering scale commutator; not leftover interpolating sequential type; not spacetime | 10.5281/zenodo.24219373 |
| Companion 51 | Three-dimensional Dushnik--Miller plateau of covering-process orders through size 10, and the unique size-11 four-dimensional witness; not leftover interpolating sequential type; not spacetime | 10.5281/zenodo.24254287 |
| Companion 52 | Dual diamond trees of same-polarity hinge complexes and vanishing covering Green of a stipulated four-vertex form; Door-1 hunt geography; not leftover interpolating sequential type; not spacetime | 10.5281/zenodo.24270436 |
| Companion 53 | Zipper pyramids, five-point stencil of a stipulated four-vertex form, and unique six-diamond chord invertibility; Door-1 hunt geography; not leftover interpolating sequential type; not spacetime | 10.5281/zenodo.24275016 |
| Companion 54 | Dual-family time and space diamonds of a stipulated four-vertex form, and exact seven-point and nine-point finite stencils; Door-1 hunt geography; not leftover interpolating sequential type; not spacetime | 10.5281/zenodo.24276320 |
| Companion 55 | Closed self-energy formulae, a unimodular causal inverse, and conserved-self-energy strips on zipper pyramids; Door-1 hunt geography; not leftover interpolating sequential type; not spacetime | 10.5281/zenodo.24282619 |
| Companion 56 | Discrete polarity-flux Bianchi identity, covering-edge volume law, and dual-family stress-energy conservation; finite covering identities, not the contracted Bianchi identity of general relativity and not a continuum field equation; Door-1 hunt geography; not leftover interpolating sequential type; not spacetime | 10.5281/zenodo.24282916 |
| Companion 57 | Krein factorization, Wilson plaquette identity, and dual-family Wronskian of a stipulated four-vertex form; finite covering identities, not a continuum Weitzenbock identity and not a continuum nonabelian gauge field equation; Door-1 hunt geography; not leftover interpolating sequential type; not spacetime | 10.5281/zenodo.24286783 |
| Author | Independent researcher | ORCID 0009-0002-4415-1209 |
The live generator is the registered kernel EK-A0007-MINIMAL-T1-T2-COVERING-V0,
read in a port-resolved convention at port budget q = 2.
Births have arity one or two.
The relative intensity of those two arities is a quantity that a selection principle would have to supply;
it is not derived from discrete general covariance alone.
On that kernel the covariant coherent dynamics, truncated at a finite horizon N, form a set affinely isomorphic to the probability simplex on the isomorphism classes of ported states. The number of those classes, by exact enumeration, is
1, 1, 2, 4, 11, 33, 118, 471, 2 126, 10 625, 58 403, 349 391
for N = 1, …, 12. At each finite-horizon extreme point the arity-two count is a class invariant, so covariance plus coherence does not pin a unique physical intensity. On the infinite growth graph the same spectrum fills the whole interval [0, 1] among constructed ergodic (tail-trivial) measures; fixing a density still leaves infinitely many pure phases. A selector, if there is one, has to import data from outside the growth graph.
The certified extremal locus used for those constructions is classified: it is exactly the admissible ladders (blockwise total orders with a singleton bottom block and no block larger than two). On that locus an independent Rideout–Sorkin p-thinning that retains the unique minimum acts as a one-parameter semigroup. Scale invariance on the locus selects the chain and nothing else. That is a statement about a flow of measures on a classified family of states. It is not a derivation of leftover r (the leftover share of arity-two births), a continuum spacetime, or a physical universe.
Proofs for this layer occupy Companion Papers 20–23. Companion Paper 24 is a related labeled-path coarse map (10.5281/zenodo.22648199); it is a different object, not a continuation of the covering-kernel classification. The geometry of the remaining mixed-extremality bar occupies Companion Paper 25, recorded above. Companions 27 and 28 (13 October 2026) record further covering-kernel facts already on deposit: order polytopes and port scaling, a finite non-identifiability pair, a bounded-valence lemma, an ordering-fraction normalisation, and the restriction that covering birth at port budget two is not Rideout–Sorkin sequential growth. Companion 29 (14 October 2026) records covering height and Dilworth width against sequential measures. Companion 30 (15 October 2026) is an exclusion theorem: bounded-valence sequential laws, including covering-kernel support, are asymptotically singular to a Minkowski sprinkling in every dimension at least two; Rideout–Sorkin classical sequential growth is independently singular to those sprinklings. Sprinkling is used as a comparator, not as a causal-set programme affiliation. Covering histories are not identified with physical spacetime. Companion 31 (15 October 2026) records the incomparability ortholattice of covering histories: the map is the standard closed-set functor of an orthogonality space; it is not a new causal logic, and it does not derive quantum logic. Companion 32 (18 October 2026) records a history-independent characterisation of covering-process orders: a finite nonempty poset is a covering history if and only if it has a unique minimum, Hasse in-degree 1 or 2 off that minimum, and Hasse out-degree at most 2. Automorphism reconstruction of maximals with isomorphic deletions is proved in named classes. Coupled reconstruction for all sizes is not claimed. Covering histories are not identified with physical spacetime. Companion 33 (20 October 2026) records sequential extra-input interpolants of covering orders versus two-dimensional sprinkling: conservative extra-input is a down-set cut; extra vertices freeze, join, or discover covering; extra diamonds fail the comparable-fraction control. No generation rule is written. Covering histories are not identified with physical spacetime. Companion 34 (20 October 2026) records handshake surplus of conservative extra-input on covering-process stems: leftover ports are vertex-local; surplus equals 2 − P + Linto − 2c; leftover-port extras are capped by unsaturated width; leftover defect chambers close locally. Caps are not later laws. Covering histories are not identified with physical spacetime. Companion 35 (23 October 2026) records convex quotients of covering-process orders: identification yields a leftover poset if and only if the subset is order-convex; leftover Hasse degrees do not increase. No generation rule is written. Covering histories are not identified with physical spacetime. Companion 36 (23 October 2026) records that covering-process evolution is to a family: interpolants need not force an instance. Covering histories are not identified with physical spacetime. Companion 37 (23 October 2026; version 2, 14 November 2026) records a unique zero-parameter path-entropy measure on covering-process orders, the all-cardinality identity of the one-step T1 envelope with U, and complete maximal-deletion reconstruction through size 10. Version 2 attaches a machine census of the unordered maximal-deletion deck at sizes 11 and 12; the size-10 reconstruction statement is unchanged. Coupled reconstruction for all sizes is not claimed. Covering histories are not identified with physical spacetime. Companion 38 (27 October 2026) records a one-parameter exponential path-entropy family on covering-process orders, sector-balanced local growth with a parameter-free co-scaling horizon, flux-thresholded Dilworth–Turan coarse quotients, and symplectic Berry holonomy of a four-invariant cochain whose degree-normalized Gram meets strict GUE spacing criteria on W8 and two-sided W9. The GUE statistic is a spectral property of the unlabelled covering Hasse graph; it is not a derivation of quantum mechanics. Joint inhabitation of R > 2 and a Lorentz-band window on a two-dimensional spectral sector grid is seed-versus-null geography, not spacetime. Covering histories are not identified with physical spacetime. Companion 39 (28 October 2026) records law-level poset limits of three named covering stepwise laws. Under t1_only the global limit is the empty kernel and every window limit is the chain kernel; neither is Lorentzian of dimension at least 2. For uniform_legal and port_resolved the remaining limit statements are conditional on a named incomparability condition, which remains a conjecture. Sequential probabilities of t1_only on unlabelled covering classes fail label-independence from size 4. Measured height log-log slopes above 1/2 at N at most 8192 are not a change of growth-law type. Covering histories are not identified with physical spacetime. This note does not claim Lorentz invariance and does not assert that covering structures cannot produce geometry. Companion 40 (2 November 2026) records the Young sector of covering growth: Young-diagram covering histories, Plancherel descent, and a Vershik–Kerov–Logan–Shepp poset-limit reading that is Level 2 under cited shape theorems. Graded chain length depends on aspect ratio, so Level 3 fails on that reading. Covering histories are not identified with physical spacetime. This note does not claim Lorentz invariance and does not assert that covering structures cannot produce geometry. Companion 42 (2 November 2026) records the Greene–Kleitman shadow and the two-spine lift of covering growth. One-box deletion produces a standard Young tableau; the map is a shadow, not a retraction onto the Young sector. Plancherel growth on Young-sector histories equals the recording tableau of labeled 1+1 sprinkling. A labeled 1+1 sprinkling corresponds to a pair of standard Young tableaux of one common shape. Companion 40, Theorem 10, remains the graded box-order reading. Covering histories are not identified with physical spacetime. This note does not claim Lorentz invariance and does not assert that covering structures cannot produce geometry. It is not a remint of Companion 36. Companion 47 (10 November 2026) records that a nonempty set of full-support central measures on the covering growth graph exists. Seed-rigid-ray mixtures fail at the unique predecessor-count-two class of size 4. A unique-maximal stem-and-tail mixture is strictly positive and is not extreme. An extreme full-support central measure is not exhibited. The construction is not Companion 27, Theorem 16, and is not Companion 37, Theorem 11. Covering histories are not identified with physical spacetime. This note does not claim Lorentz invariance. Companion 48 (10 November 2026) records that unique-token extras and the ultrahomogeneous embedding squeeze are not leftover interpolating sequential type. Consult independence is not leftover interpolant independence. The Aut-quotient of covering-class embeddings into the Fraïssé limit recovers unlabelled covering classes. Kechris–Pestov–Todorčević is cited, not re-proved. Covering histories are not identified with physical spacetime. This note does not claim Lorentz invariance. Companion 49 (10 November 2026) records covering-native operator geography of the covering-law family. Leftover interpolant independence failure of covering-law functors is occupancy geography, not a ban on covering-kernel papers. Scale-commutator inertia on maximal deletion with port-resolved is n_-=1, 2, 4 at sizes 6, 7 and 8; binary doubling is a look, not a theorem. A unique W_7 outdegree class is the paired-join covering of Companion 36, Theorem 13. Covering histories are not identified with physical spacetime. This note does not claim Lorentz invariance. Companion 50 (11 November 2026; version 2, 14 November 2026) records the all-n column-stochastic left identity of the covering scale commutator, the four-size Lorentzian plateau of S, complementary negative directions under T, and the flux matrix F as distinct from S. Version 2 attaches a size-10 flux census; Theorem 10 remains the five-size window and is not extended. Covering histories are not identified with physical spacetime. This note does not claim Lorentz invariance. Companion 51 (15 November 2026; version 2, 15 November 2026; version 3, 16 November 2026) records a complete SAT census of Dushnik--Miller dimension on covering-process orders: sizes 8, 9 and 10 have exact dimension 3 on every class of dimension at least 3, and size 11 has a unique dimension-4 class. Version 2 attaches a capacity certificate for that unique class: every linear extension reverses at most four of the twelve critical pairs, and no triple of capacity-4 extensions covers all twelve. Version 3 names an explicit four-realizer of that unique class. Companion 34 Theorem 23 is cited; the plateau is not a corollary of saturated width. Covering histories are not identified with physical spacetime. This note does not claim Lorentz invariance. Companion 52 (16 November 2026; version 2, 16 November 2026; version 3, 16 November 2026) records dual diamond trees of same-polarity hinge complexes, vanishing covering Green of a stipulated four-vertex form, a hinge census through eight diamonds, mixed nested-cut polarity, and a two-edge wedge census through seven diamonds whose apex graph is a tree. Version 3 attaches a first-order waist-root cap inertia addendum of assembled L_K. Occupying leftover interpolating sequential type by those objects is too loose. Covering histories are not identified with physical spacetime. This note does not claim Lorentz invariance. Companion 53 (16 November 2026) records zipper pyramids of a stipulated four-vertex form, a five-point interior stencil, unique six-diamond chord invertibility, unique self-energy and bosonic-energy minima on the height-four zipper, and V-cap parity counts. Occupying leftover interpolating sequential type by those objects is too loose. Covering histories are not identified with physical spacetime. This note does not claim Lorentz invariance. Companion 54 (16 November 2026) records dual-family time and space diamonds of a stipulated four-vertex form, covering multiplicity at most two, an exact seven-point interior stencil on the (2+1) window of side 4, and an exact nine-point interior stencil on the (3+1) window of side 4. Occupying leftover interpolating sequential type by those objects is too loose. Covering histories are not identified with physical spacetime. This note does not claim Lorentz invariance. Companion 55 (16 November 2026) records closed self-energy formulae of zipper pyramids, a unimodular causal inverse, conserved-self-energy strips, and a scoped traveling-wave identity. Occupying leftover interpolating sequential type by those objects is too loose. Covering histories are not identified with physical spacetime. This note does not claim Lorentz invariance. Companion 56 (16 November 2026) records a discrete polarity-flux Bianchi identity of covering edges, a covering-edge volume law, and dual-family discrete stress-energy conservation on the Companion 54 stencil in spatial dimension 2 and 3. The polarity-flux identity is not the contracted Bianchi identity of general relativity. The conservation identity is a discrete stencil identity, not a continuum field equation. Occupying leftover interpolating sequential type by those objects is too loose. Covering histories are not identified with physical spacetime. This note does not claim Lorentz invariance. Companion 57 (16 November 2026) records a Krein factorization of even and odd blocks of a stipulated four-vertex assembly, a Wilson four-cycle identity when covering multiplicities take only the values 1 and 2, and a dual-family bilinear Wronskian in assembled spatial dimension 1, 2 and 3. The native one-spatial-direction stress-energy identity holds in dimension 1; the planar two-direction formulae of Companion 56 Theorem 4 continue to fail on that construction. The Krein factorization is not a continuum Weitzenbock identity. The Wilson split is not a continuum nonabelian gauge field equation. Occupying leftover interpolating sequential type by those objects is too loose. Covering histories are not identified with physical spacetime. This note does not claim Lorentz invariance. They are further children of the parent, not new versions of Companions 20–56. Companion 41 is not on this table. They do not supply a later interpolating constructor, and they do not construct spacetime. Spectral algebra, negatives, the form note, and the independent manuscript that are not this kernel occupy their own class pages. The parent manuscript is the Quantum Narrative Matrix / covering-kernel text on Zenodo.
Please judge the deposits, the named unpaid objects, and the claim boundary on each page.
Companions 20–23 exhibit covariant coherent dynamics on the kernel. They do not exhibit a point of covering K.
A further layer, occupied in Companion Paper 25 (15 September 2026), records the already-deposited geometry of the remaining mixed-extremality bar, without claiming that bar paid. Three compacta on this axis are not the same set: the inverse-limit covering compactum of the covariant simplices; the covering-face compactum on which mixed extremality is posed; and the every-layer leaf Meyer box last ≤ 3/4. Occupancy of a Meyer-range slice is not mixed extremality of covering K.
The deposited mixed-extremality clause is a liminf: exhibit one point of the covering-face compactum whose covering birth-density sequence satisfies lim inf Fn ≤ 3/4. Fatou’s lemma reduces a Choquet clamp at threshold 1 to that inequality at threshold 3/4 without exhibiting a point. Co-density converts the box inequality into a budget that Dirac reverse-equal complete-parent edges cannot carry. An exact census through event-count 11 is on record.
A point of the covering-face compactum is exactly an all-N reverse-equal compatible family. Event-count 2 is structurally over-wall, so the last-cut question begins at event-count 3. An exact downward last-cut profile records 923 of 58 403 classes at event-count 11 remaining in the box at every event-count from 3 through 11. The associated witness polytopes are nonempty at event-counts 9, 10 and 11. Unique-minimum of covering classes is a covering-fire invariant and is not a characterisation of those polytopes. Nonemptiness of the witness polytope at every event-count remains open. That all-M statement is mixed extremality; it is not claimed here. Companion Paper 25 is a further child of the parent, not a new version of Companions 20–24. This child is 10.5281/zenodo.22769118. Companion Paper 26 occupies the inhabitation of that deposited liminf clause by a two-chain last closed form, and records that a Fatou extreme with that liminf lies outside the closed convex hull of vertex threads. It is a further child of the parent, not a new version of Companion 25. This child is 10.5281/zenodo.22919449.
Sequential growth is already a recognised way to ask how a discrete causal order can be generated rather than assumed. What this kernel adds is a small, registered generator on which several questions can be answered by exact classification instead of by slogan: which covariant laws exist at a finite horizon; that covariance and tail-triviality do not select an intensity; what the certified extremal states actually look like; and how a standard thinning flow acts on that locus.
Those no-go results are useful because they say where not to look. The remaining work is then forced to be honest: any physical selection has to come from an object that is not already implied by the growth graph, and mixed extremality of covering K has to be an inverse-limit point of the covering-face compactum, not a finite prefix and not occupancy of a Meyer-range slice. That is a narrower, checkable problem. I think the direction is worth continuing for that reason — the questions are well posed, the certificates are machine-readable, and the unpaid mouths are named rather than hidden.
If later work pays a named mouth, that should be recorded as a scoped result. Until then, the correct public sentence is the one above: the foundational work of this direction is on record; cosmogenesis is not.