Papers ← Paper A2 · Braided ← Chapter A3A
Paper A3A — Creation: The Algebra
The Clifford algebra forced by Paper A0’s inner-product space — the positive-definite Cl(3,0) — and its descent to the birth of time, where the fourth generator e₀ enters and the algebra becomes the Lorentzian Cl(3,1). The seven non-scalar elements, the two crossings and the provable asymmetry between them, the constraint cascade and its doubling, the Dirac equation, and E = mc².
Confidence — Math: derivation at the spine — the algebra, the two crossings, the cascade and its levels and their nesting, the Dirac form, c’s invariance — from four named inputs, the cascade’s three posits (P1–P3) and §3’s constitution. Around it stand five identifications, four inferences (§3’s three constraints; the rate law; time’s arrow; antimatter’s absence above the crossing) and one further posit (the access principle). Two results carry a test and the test is A3C’s; one stands open, and the rest is refutable by error, not by observation. Science: — (not engaged) — the forms are derived at §6; the measured universe is A3C’s. Theology: — (not engaged). For per-result ground, support and exposure, see the Traceable Account at the foot.
Abstract
In A0’s space, states could be measured against one another but not yet made to act on one another: A0 supplied a scalar-valued inner product, but no vector-valued multiplication — no way for one state to operate upon another. That multiplication is derived below, and shown forced — given one postulate: that a state’s square is its existence-verdict, a scalar of the metric. Under that postulate the product can be nothing but the Clifford product. Three further inputs carry the result the rest of the way: the cogito (A1) — that this exists, concluded by existing, the chain’s one self-attesting given — Ehrenfest stability, and the complement posit (P0, §6). Together they force the algebra of the created order to be Cl(3,1). Its positive-definite part, the one A0’s metric alone writes down, is Cl(3,0): the e₀-free spatial half of the whole. The three directions are the world’s own, and Ehrenfest stability shows they could not have been more — no stable orbit or atom survives four or more spatial dimensions; at three, one algebra answers and the other splits, which is what settles the sign of the defining relation (§1). The Lorentzian −1 is then fixed by elimination (§6).
The rest descends from Cl(3,0), the spatial part. It has seven non-scalar elements across three grades, the one product read at each grade, and the vector↔bivector duality special to three dimensions; §2’s asymmetry then fixes exactly two qualitative crossings and the provable asymmetry between them (grade 2 reachable from grade-1 means, grade 3 not from grade-2 means). The descent proper follows. A single constraint turns out to require three structures at once — time, gravity, spacetime. From there: the constraint cascade and why it doubles, and the dimensional hierarchy it produces. Last comes the crossing that is the birth of time: e₀ enters, the algebra in force becomes Cl(3,1), and with it arrive the Lorentzian signature, the Dirac equation, E = mc², and the speed of light as a structural output rather than an input.
“Eternity is in love with the productions of time.” — William Blake, The Marriage of Heaven and Hell (c. 1790)
Introduction
Paper A0 gave a space and a way to measure within it: a complex inner-product space, in which any two states can be compared — ⟨v, w⟩ returns a number, the extent to which one overlaps the other. With it came the theorem by which that space is actualized: the Generalized Nuclear Spectral Theorem (GNST) (A0 §3.3) — which this paper reads in its operative, successive sense, the act by which distributional potential is resolved into definite states, event by event. What A0 does not yet carry is a product: its inner product takes two vectors to a scalar; nothing in A0 takes two vectors to another vector — the space can be measured, but its directions cannot yet act on one another. That next operation — multiplication — is where this paper continues.
An algebra — a vector space equipped with a multiplication (Mathematical Preliminaries, below) — turns the space of states into a space of operations: directions that compose, rotate, and act, not merely are measured. A product is what actualization requires — but the requirement has two faces, and they must not be run together.
Within a state, a product reads its verdict. To bring a state from may or may not be into is or is not is to let the state be taken against itself, and the reading is a magnitude: v² = ⟨v,v⟩ · 1. Nothing is carried anywhere; what the square returns is this state’s own standing. This is the actualization postulate, motivated in §1 — and it is a posit, not a theorem: it is what the GNST asks of the world’s directions, and motivation is not proof of exclusivity.
Across states, a product carries. ψ to φ, one actual thing to the next. That is acting in the ordinary sense, and it is what succession is made of (§6). It is a different demand on the product than the verdict, and a state that could only be squared could never go anywhere.
One product answers both, which is why only one is needed: its symmetric grade is the measure that returns the verdict, its antisymmetric grade the plane in which anything can be carried at all (The one product, two grades, §1). The distinction is not decorative. A level carrying grade 0 alone has verdicts and no movement — being, actualized, with nothing carried anywhere — which is exactly what the apex is (A3B §H₁).
The Cascade begins with a single determination: which multiplication A0’s space carries.
The answer is forced, up to one postulate §1 names and motivates. Given A0 — its positive-definite metric, its associative domain, its complex state space — and one further given, already standing in the chain: the cogito (A1, Assumption II), the certainty whose denial performs it. That this exists is concluded by existing; and whatever concludes persists through its concluding, so a persistent thing — the knower, a state ψ ∈ H — comes attested with it. On these — the postulate granted — the multiplication can only be the Clifford product. Three inputs carry the result the rest of the way — the cogito already named above, and two more, which §7 enumerates with the rest: Ehrenfest’s stability, constraining the spatial count while leaving a three-dimensional complement open; and the complement posit (P0, §6), without which there is no fourth generator to have. Beside them stands one identification, which §7 holds apart from that list — i ≡ I, by which the algebra reproduces A0’s complex state space, and on which the count does not depend. With those, the algebra of the created order can only be Cl(3,1), of which Cl(3,0) — the positive-definite part A0’s metric alone writes down — is the e₀-free spatial part. The count itself is not among the inputs: §1 takes it from the world the cogito commits the chain to, and shows it could have been no other. From there the whole descent follows.
Mathematical Preliminaries
Algebras
An algebra over a field K (here ℝ or ℂ) is a vector space A equipped with a multiplication A × A → A that is bilinear — distributing over addition and commuting with scalars — so that the product of two elements is again an element of A. It is associative when (ab)c = a(bc). A0 supplied an inner product, which is bilinear but scalar-valued (V × V → K); an algebra is the further structure in which vectors multiply to vectors, and so can act on one another (Lang, Algebra; Bourbaki, Algebra I).
Clifford algebras
Let V be a real vector space carrying a metric — a symmetric bilinear form ⟨·,·⟩, A0’s inner product. The Clifford algebra Cl(V, ⟨·,·⟩) is the associative algebra generated by the vectors of V subject to the single relation
v² = ⟨v, v⟩ · 1 for every v ∈ V,
equivalently, on an orthonormal basis {e₁, …, eₙ}, eᵢ² = ⟨eᵢ, eᵢ⟩ and eᵢeⱼ = −eⱼeᵢ for i ≠ j — distinct generators anticommute. This is the defining move: the metric is carried inside the multiplication, so that squaring a vector evaluates its length. (Set the right-hand side to zero instead, v² = 0, and one obtains the exterior algebra, which discards the metric; actualization forbids that choice (§1).)
The n vectors e₁, …, eₙ are the algebra’s generators, and n is the dimension of the space they span — the sense in which “dimension” is used throughout when the number of generators is at issue. The algebra built on them has a dimension of its own, 2ⁿ, which is a different number and is never what the count argument refers to. Products of distinct generators sort the algebra into grades: grade 0 the scalars, grade 1 the vectors, grade 2 the bivectors (eᵢeⱼ, i < j), up to grade n, the single pseudoscalar e₁⋯eₙ. A Clifford algebra on n generators is 2ⁿ-dimensional. (A note on terms: because v² = ⟨v,v⟩ can lower degree, a Clifford algebra is filtered rather than Z-graded — its associated graded algebra is the exterior algebra — and only the Z₂-grading into even and odd parts is exact. The grade decomposition used throughout is a decomposition of the underlying vector space, in the standard geometric-algebra idiom; where an argument turns on the grading being closed under multiplication (§2, §6), it is the Z₂-grading doing the work. Lawson & Michelsohn I.1.) The product of two vectors splits into a symmetric (grade-0) and an antisymmetric (grade-2) part,
ab = ⟨a, b⟩ + a ∧ b,
with ⟨a, b⟩ = ½(ab + ba) the scalar and a ∧ b = ½(ab − ba) the bivector; this is the geometric product (Hestenes & Sobczyk; Lounesto). Signature is recorded as Cl(p, q): p generators squaring to +1, q to −1.
One choice is made here and is not a convenience. The defining relation is written two ways in the literature — v² = +⟨v,v⟩ (Hestenes; Lounesto) and v² = −⟨v,v⟩ (Lawson & Michelsohn) — and two quite different things follow from that, which must be kept apart. One is bookkeeping: the label Cl(p, q) denotes different algebras in the two traditions, so a given algebra carries opposite labels in their tables, and a citation must be read in its own convention.
The other is not bookkeeping at all. Applied to one and the same positive-definite metric, the two relations build genuinely different algebras — at three directions, M₂(ℂ) under the first and ℍ ⊕ ℍ under the second — so which relation is written is a claim about the world and not a choice of notation.
This paper takes the first, and §1 earns it rather than adopting it. The argument runs in that order: the number of directions is fixed first, by exclusions that hold in both conventions alike, and only then the sign — settled by which of the two algebras at three directions is a world, the split one being neither one world nor able to carry a phase. The sign is therefore an output of §1, not an input to it. Until then the +1s here are provisional labels.
“It is wrong always, everywhere, and for anyone, to believe anything upon insufficient evidence.” — William Kingdon Clifford, “The Ethics of Belief” (1877)
Simple algebras, matrix algebras, and spinors
An algebra is simple if it has no proper two-sided ideal — no nontrivial part closed under multiplication by the whole. A simple algebra cannot be split apart, and has, up to isomorphism, exactly one irreducible representation — one essential way of acting on a vector space, its spinor module. The classification of the Clifford algebras (Bott periodicity) identifies each Cl(n, 0) with a matrix algebra over ℝ, ℂ, or ℍ; in particular Cl(3, 0) ≅ M₂(ℂ), the 2×2 complex matrices, whose spinor module is ℂ² — the two-level complex state, on which the three generators act as the Pauli matrices (Lawson & Michelsohn; Lounesto). These are the facts that single out Cl(3, 0) in §1.
1. The algebra
The positive-definite algebra A0’s inner product carries is Cl(3,0) — the e₀-free spatial part of the created order’s full algebra, Cl(3,1) (developed below). A0 fixes the kind, in two respects. The metric is positive-definite — A0’s inner product has ⟨v,v⟩ ≥ 0, zero only at v = 0, which is A0’s metric absolute. And the algebra is associative, because A0’s nuclear domain is complete self-consistency under every probe: group the acts of measurement however you like and the nesting never changes the answer. An algebra that reassociates to a different product, as the octonions do, cannot honor that invariance — so the algebra of creation cannot be one.
The real and the complex. Two structures must be kept straight, and each is canonically supplied. The Clifford construction runs on a real vector space carrying a symmetric bilinear form, and A0’s Hermitian inner product provides one without further assumption. Every complex Hilbert space is a real inner-product space under Re⟨·,·⟩, and on the span of the generators that real part is the metric the relation v² = ⟨v,v⟩ carries inside the product. As the cascade deploys them at H₃ (§4) the generators are created directions — a real three-span bearing Re⟨·,·⟩, positive-definite because A0’s metric is. The complex structure of H is then not presupposed by the algebra but required of it: an algebra of creation must reproduce the complex state space A0’s states live in, manufacturing a central i of its own. What plays that i is the pseudoscalar I = e₁e₂e₃ (central, I² = −1): the identification i ≡ I (Hestenes). It is an identification, not a theorem — and its vindication is that the algebra it selects hands back exactly A0’s complex two-level state as its spinor module ℂ² (Preliminaries): complexity goes in as a requirement and comes out as an output.
Why a Clifford algebra, and not the metric-blind one. The selection runs in three steps — two posits and a theorem, and the word forced belongs to the theorem. (i) The scalar verdict. To actualize is to turn a may or may not be into an is or is not — so the algebra must read each state’s verdict off the state itself, and the reading is its square. A verdict of existence is a magnitude — a number, not a structured element — so v² must be scalar. (ii) Whose magnitude. The only measure A0 supplies is its inner product, so that scalar can only be the state’s own length: v² = ±⟨v,v⟩ · 1 — what the postulate fixes is that the square is the metric’s scalar, and nothing more; the sign is settled below. Together (i) and (ii) are the actualization postulate — motivated by what the GNST asks (the world’s directions must act, be operative, invertible elements rather than inert labels), though motivation is not proof of exclusivity. (iii) The theorem. An associative algebra generated by the space’s directions subject to v² = ⟨v,v⟩ is, by the universal property, the Clifford algebra — here, and only here, forced is the exact word. What the postulate buys, and its denial costs, is read straight off v²:
- Existence. With A0’s metric carried, v² is nonzero for every nonzero vector — every actualized state genuinely is, and only the zero state is not. That is exactly the is/is-not dichotomy actualization produces, and what produces it is the square’s vanishing or not, never its sign. The exterior algebra sets v² = 0 for every vector: all “is not,” nothing actual — the algebra of pure potentiality, in which no state is ever created.
- Invertibility. In the Clifford algebra a vector has an inverse, v⁻¹ = v/⟨v,v⟩, because v·v = ⟨v,v⟩ ≠ 0 — so vectors can act, compose, rotate (the whole Spin/rotation machinery lives here). In the exterior algebra v² = 0 makes every vector a zero-divisor — non-invertible, inert. Nothing can act.
So “algebra of creation” and “Clifford algebra” are one requirement, joined through the postulate: the exterior algebra — the metric-blind choice, v² = 0, the postulate’s denial — is the one in which nothing is actual and nothing acts, and a creation staged there would produce only the null state. To create at all is to give a state a magnitude that does not vanish — which is what the Clifford relation provides. The postulate motivated, the theorem does the rest.
That leaves the dimension — the number of generators. A generator is a direction of the space, so the number of generators is its dimension, and the algebra’s dimension and the world’s are one number. It is therefore not settled by a search for the smallest algebra that would do. The algebra of creation must model the creation there is — this world, the one the chain is already committed to by the cogito (A1, Assumption II): a knower exists, holds a state ψ ∈ H, and persists through his own knowing. That world has three directions, and what follows shows that it could have had no other number, every exclusion below running in both sign conventions alike. Only then does the sign the Preliminaries left open become answerable, and it is answered by what survives at three.
At none, the algebra is ℝ: scalars alone, no directions, no things — magnitude with nothing to bear it. This is what the cascade requires at H₁, grade 0 alone, ⟨·,·⟩ and nothing else. What in fact obtains there is past reach — the apex is “invisible in principle to a creature within the created order” (A3B) — so this names what the descent needs above it and claims no sight of it. The level itself is not empty: least constrained, H₁ is the largest space, containing every level beneath it (§4).
At one direction, the algebra is commutative whichever sign is taken: ℝ ⊕ ℝ when the generator squares to +1, ℂ when it squares to −1. The generator is invertible in both — that much the postulate secures, and it is not what fails here. What fails is that there is nothing for it to act on: conjugation is trivial in a commutative algebra, vxv⁻¹ = x for every x, and with a single direction there is no bivector, hence no plane, no rotation, and no second direction to be carried into. A world of one direction is not a thin world but one in which no thing bears on any other.
At two directions, the first exclusion is the one this section opened with: the algebra of creation must model the creation there is, its dimension and the world’s being one number — and this world is not two-dimensional. Ehrenfest does not reach the case (two-dimensional atoms bind perfectly well), but it does not need to; a stable two-dimensional world is still not the one the cogito is committed to. A second and independent exclusion holds even of a reader who grants nothing about how many directions there are. The centre is ℝ in both conventions, for a reason of parity rather than of sign: whenever the number of generators is even the pseudoscalar anticommutes with the odd part and so is not central, and nothing else is either. So neither Cl(2,0) ≅ M₂(ℝ) nor Cl(0,2) ≅ ℍ holds an i commuting with everything — different algebras, the same failure — and neither can carry a consistent global phase or a complex state. The test is not that A0’s H is complex and these fail to contain it; it is that the algebra of creation must reproduce the complex state space A0 requires (The real and the complex, above), and neither can manufacture the central i that takes. Two directions give a world of real quantities, not one in which quantum things — states with phase, superposition — exist at all.
The scope of that test. It is asked of the algebra A0’s metric writes down directly — the algebra of the state-bearing directions, nothing else yet in it — where the states are the algebra’s own irreducible module, so the imaginary they need must commute with all of it. It stops being the right question the moment a direction enters that bears none: when the complement e₀ arrives at the crossing (§6), the created order’s algebra is Cl(3,1), whose centre is also ℝ, and the complexity has moved into the even part, where the states then live. So the parity fact above governs this step and is not a standing verdict on every algebra with an even number of generators.
Above three the exclusions so far say nothing, and the candidates there are not empty. Cl(7,0) ≅ M₈(ℂ), Cl(11,0), and on up are single and complex too, the single, complex members of this family being the Cl(n,0) with n ≡ 3 mod 4. What rules them out is not that they are larger. It is the same knower.
From four directions up the binding force falls as 1/r³ or faster, so the potential goes as 1/r² or steeper and no stable bound state survives. At four the potential is exactly the scale-invariant critical case, and the two sides of its threshold fail differently. Below the critical coupling the spectrum is continuous and holds nothing bound at all, while above it the Hamiltonian is unbounded below, the levels accumulating to −∞ with no ground state, which is the fall to the centre. From five up it is unbounded below outright. Either way nothing persists — which disposes of the higher single-and-complex members and of everything between them alike. That is the Ehrenfest stability result.
Note what it imports — not bare “stability” but the shape of the binding law, Gauss-law falloff and the quantum Hamiltonian it bounds; the physics that descends in §6 is therefore fixed relative to this anchor.
Such a world holds no stable structure — nothing that can persist as a thing; and something does persist, the knower who endures his own concluding. A world with no persistent thing cannot be this one.
(Nor can the extra dimensions quietly curl away: compactification is a mechanism, and the framework supplies none — the generators enter symmetrically, nothing in the algebra distinguishes a direction to shrink, so a curled higher world would need machinery imported from outside everything derived here.) So the count is closed on every side: at one direction nothing acts on anything, at two nothing can carry a phase, and from four up nothing persists. Three is what remains — and three is what the world has.
Which leaves the family, and with it the sign. The count fixed, two algebras answer to it — Cl(3,0) under v² = +⟨v,v⟩ and Cl(0,3) under v² = −⟨v,v⟩ (Preliminaries) — and only one of them is a world. And the whole difference is one square.
Three being odd, the centre is spanned by 1 and the pseudoscalar ω, so what the centre is depends entirely on ω². Under v² = +⟨v,v⟩ that square is −1: the centre is ℂ, the algebra is simple, and Cl(3,0) ≅ M₂(ℂ) — one world, complex, spinor module ℂ², exactly the two-level complex quantum state, and already complete, M₂(ℂ) being the full quantum system. Under v² = −⟨v,v⟩ it is +1: the idempotents (1 ± ω)/2 split the algebra into Cl(0,3) ≅ ℍ ⊕ ℍ, two halves with no traffic between them, whose centre is ℝ ⊕ ℝ and holds no imaginary. (The algebra itself is full of square roots of −1, being quaternionic; what it lacks is one that commutes with everything — the same lack that disqualified two directions.)
So one fact carries two consequences, and either one alone disqualifies the negative sign: ω² = +1 both splits the algebra and denies its centre an imaginary. The two are not independent facts, but they are independent grounds — a world that is two halves with no traffic between them fails whether or not anything is said about phase, and this matters for the accounting, because the phase requirement reaches the algebra through an identification (i ≡ I) and the splitting does not.
At three directions, then, the sign rests on a ground that needs no identification — where at two directions the simplicity route was closed, Cl(2,0) being simple, and the exclusion ran instead by the world-premise this section opened with, with the phase argument standing independently beside it. So the sign of the defining relation is neither adopted nor a convenience; it is what is left standing at the count the world has.
So the algebra of creation is Cl(3,0), its symmetric grade A0’s own inner product and everything above it following from A0’s actualization. (The spatial directions are realized as physical three-space — the relative vectors σₖ = eₖe₀ — at H₄₈.)
The same result comes from inside the algebra. Positive-definite squaring, applied to e₁ + e₂, gives (e₁ + e₂)² = 2, which forces e₁e₂ + e₂e₁ = 0: the generators must anticommute. And an associative algebra on three anticommuting generators, each squaring to +1, is the universal Clifford algebra. Cl(3,0) is moreover simple, admitting no proper quotient, so the uniqueness is complete. Its single irreducible representation is the spinor S ≅ ℂ² — one kind of instance, which the descent then multiplies without ever varying. Given the three, Cl(3,0) determines everything downstream: the seven non-scalar elements, the grade structure, the constraint cascade, the two qualitative crossings. Associativity is the condition under which determinate identity is possible at all: a non-associative creation would make “what a thing is” depend on the order its operators were evaluated — truth itself evaluation-order-dependent.
The one product, two grades. The defining relation does more than select the algebra — it fixes what a single multiplication is. For any two vectors a and b, the geometric product splits once and completely: ab = ⟨a,b⟩ + a∧b, its symmetric part ⟨a,b⟩ = ½(ab + ba) a scalar, its antisymmetric part a∧b = ½(ab − ba) a bivector. This is not a second structure laid over the algebra; it is the one product read at its two grades — and the derivation just given is its first instance. (e₁ + e₂)² = 2 is precisely the symmetric part evaluated, and the forced anticommutation e₁e₂ + e₂e₁ = 0 is that symmetric part vanishing in the case ⟨e₁,e₂⟩ = 0. So the inner product is not a separate condition standing outside the algebra and constraining it — it is the algebra’s symmetric grade, and the anticommutation is the ⟨a,b⟩ = 0 case of a universal split. The symmetric grade is the metric — the relation that measures; the antisymmetric grade is the paired mode — the bivector, the plane spanned by the two vectors it joins (The pseudoscalar duality, below). Measuring and pairing are the symmetric and antisymmetric faces of a single act: one product, read from two sides.
The seven non-scalar elements. Cl(3,0) is eight-dimensional: one basis element is the scalar, and the other seven (2³ − 1 = 7) are non-scalar, qualitatively distinct by algebraic character:
| Element | Grade | Squares to |
|---|---|---|
| e₁ / e₂ / e₃ (vectors) | 1 | +1 |
| e₁e₂ / e₁e₃ / e₂e₃ (bivectors) | 2 | −1 |
| e₁e₂e₃ (pseudoscalar) | 3 | −1 |
Grade-1 elements square to +1; grade-2 and grade-3 elements square to −1. The −1 is what makes an element generate rotation: for a bivector B with B² = −1, exp(Bθ) = cos θ + B sin θ, a rotation in the plane B spans; for the central pseudoscalar it is the global phase on ℂ² (i ≡ I, above). That is a fact about the exponential — X² = −1 gives the circular functions where X² = +1 gives the hyperbolic.
The pseudoscalar duality. In three generators there is a further identity the table conceals: each bivector is the pseudoscalar-dual of the complementary vector. With I = e₁e₂e₃ (the pseudoscalar, I² = −1, central), the bivectors satisfy e₁e₂ = I·e₃, e₂e₃ = I·e₁, e₃e₁ = I·e₂ — so a pairing of two vectors and the third vector standing alone are one element wearing two descriptions. And it is special to three: a bivector is dual to a vector only in three dimensions (in n dimensions it is dual to an (n−2)-vector), so the coincidence holds precisely because the generators are three.
Cl(3,0) is the positive-definite part; the whole is Cl(3,1). A0’s metric alone reaches the state-bearing directions and no further, so what it can force directly is the algebra they generate — Cl(3,0), by the count and then the sign, above. But the metric is not the only given. The cogito’s — that this exists, with a persistent knower in it — reaches further, and fixes the fourth generator by elimination (§6). e₀² can be neither +1, a fourth spatial dimension and Ehrenfest-unstable, nor 0, a null and inert time — a direction that cannot act, so cannot carry the succession of acts the GNST performs through it. So e₀² = −1. So the algebra of the created order is Cl(3,1), and Cl(3,0) is its e₀-free spatial part, deployed first and in order by the GNST — generators, bivectors, pseudoscalar — before the e₀-involving part enters at the birth of time (§6).
2. The two crossings
Cl(3,0) has four grades of dimension 1, 3, 3, 1. Of the three transitions between them, the first — grade 0 to grade 1, which is the deployment H₁ → H₃ — is not a selection at all but constitution (§4), so it is not a crossing of this kind. Of the remaining two, both are qualitative crossings — qualitative in §1’s sense, the arriving class differing in algebraic character from every class already deployed: how its elements square, whether they commute, what they do — and they differ in a formally provable respect — one is bridgeable by its predecessor’s own means, the other is not. (Neither is a genuine gap in the algebra: both resources lie inside Cl(3,0); the one input no element provides belongs to the further crossing below, H₁₂ → H₂₄.)
The grade-1 → grade-2 crossing is bridgeable from within. A bivector is the product of two distinct generators (eᵢeⱼ, i ≠ j); no generator reaches grade 2 alone, since eᵢ² = +1 returns the scalar. The crossing requires combination — the +1 generators giving way to bivectors that square to −1, the first appearance of rotation — but the resource that bridges it is already present: a second generator, outside any single-generator subprocess yet inside the algebra. It draws a new paired mode from parts already given.
The grade-2 → grade-3 crossing is not bridgeable by grade-2 means at all. Every Clifford algebra carries the Z₂-grading — the split into even and odd parts fixed by the grade involution, under which even × even = even, even × odd = odd, and odd × odd = even (Lounesto §3; Lawson & Michelsohn I.1). The even part is therefore a subalgebra and the odd part is not; that is what graded means. The consequence the descent needs: every product and sum of grade-2 elements stays inside the even subalgebra Cl⁰(3,0) ≅ ℍ — grades 0 and 2 only — so no recombination of paired modes, however extended, ever reaches grade 3. Only an odd element — a generator entering the product — gets there. That is the provable asymmetry between the two crossings: grade 2 is reachable by grade-1 means, grade 3 is not reachable by grade-2 means. No recombination of the current mode attains it; only an odd element does — and its resource, the third generator, lies inside the algebra, not outside it. The one input no element of the algebra provides belongs to the further crossing below (H₁₂ → H₂₄).
The asymmetry is the algebra’s parity — even products never leave the even part — and its role in the descent is exact: parity is what makes latency possible, an element folded into every product yet unreachable by the deployed stratum’s own means, appearing only when an odd element is admitted (§6, the e₀ latency).
Of these two crossings, the second — grade 2 → grade 3 — forms the pseudoscalar: the whole, the product of all three generators, present in the algebra once all three are. There is a further crossing beyond both, lying past the algebra rather than within it: the exit, H₁₂ → H₂₄, where the generator that enters is none of the algebra’s own elements — no product of e₁, e₂, e₃ yields it. That — not the grade-2 → grade-3 crossing — is the one input no element of the algebra provides; that it is nonetheless entailed — cast by the selection as its complement — is §6’s result.
3. The single constraint requires three structures
The spectral theorem describes how potential collapses to a definite outcome; it does not specify where the center of the space is or what its topology is. Two eternal structures supply what is missing — eternal throughout this paper meaning not produced by the GNST, but standing prior to the succession they make possible — and they are the structural grounds of two forces:
- The norm ‖·‖ is what gravitates. The norm is the non-relational face of the inner product — the absolute magnitude of each state taken alone: A0’s metric absolute, “the center of gravity of the space” (A0 §5.3). The metric is the algebra’s own symmetric grade — eᵢeⱼ + eⱼeᵢ = 2gᵢⱼ, carried by the product itself (§1, The one product, two grades) — and gravity is the variation of that metric from place to place. The norm is what sources the variation: where general relativity writes T_μν, this framework reads the norm registering actualized content. Its physics is developed in A3C.
- Spacetime is the nuclear topology τ_nuclear. The spectral measure dμ is not well-defined without a topology — convergence, continuity, limits all require one; the nuclear topology supplies it. The identification reads spacetime as what that topology looks like from inside the created field — the structure fixing which events are adjacent and which paths continuous — the framework’s reading, on the same tier.
So the complete specification at the first created level is three constraints: GNST (time), the norm (gravity), the nuclear topology (spacetime). The step from one constraint to three is constitution, not selection — the completion of the mechanism’s description.
4. The cascade and why it doubles
Below the constitutional step, the hierarchy descends by eigenstate selection. Each step is the spectral mechanism collapsing a distributional potential to one outcome. This is A0’s GNST read in its operative sense — actualization as a successive, discrete act: the world reconstituted event by event, moment to moment, rather than sustained in one continuous flow.1
Why the count doubles depends on keeping two counts apart. The algebra fixes the structure — the kinds of element, grade by grade (1, 3, 3, 1): the alphabet of what can exist, and nothing about how many determinate states are spelled from it. The cascade count is the other quantity — the constraints in force at each level, which is what n indexes: the words, not the letters. (What doubling them doubles is the multiplicity of determinate states; The representation tower, below, makes that exact.) To the alphabet a single distinction adds one kind — a grade-class, deployed whole, as one act; a class may carry one element or three, its threeness the frame’s, inherited from the three directions, never three separate collapses. To the count it is a doubling, since every state the constraints already fix now comes in two versions, this or that — and because the constitutional triple is indivisible (§3), each version carries all three, so the constraints in force double with them. A GNST collapse is one such distinction — at its simplest a yes/no, a single bit — so each rung doubles the determinate states.
Handedness is the case one can see whole: the pseudoscalar deploying at H₁₂ adds one kind to the structure — the volume I. Fixing its sign fixes an orientation — a right frame or a left — and that single choice distinguishes states below it that were previously indistinguishable. One kind; twice the words. The rungs that deploy a threefold class (the bivectors at H₆, the σₖ at H₄₈, the trivectors at H₉₆) are the same single act: the class enters whole. Were each element its own collapse, the count would multiply by eight at those rungs rather than double; P3 is what excludes that, and the doubling is its consequence.
The constraint count doubles at every step, and the doubling is carried by information. n counts the constraints in force, and the three ride together: §3’s triple is indivisible — the operative GNST requires the norm to fix magnitude and the nuclear topology to make dμ well-defined at all — so wherever the mechanism operates, all three operate. A single binary distinction gives everything already fixed two versions, and each version carries the whole triple. A bit does not lengthen the rulebook by a line; it doubles the places the same three rules are in force — which is why the count doubles rather than increments, and why it doubles by two rather than by three. The derivation runs on three posits, named: (P1) each level-transition is a single GNST collapse; (P2) a collapse is a single binary distinction — that a difference, at bottom, is a yes/no, which is very near what a distinction is; (P3) each collapse deploys the next grade-class whole, as one act (above). On those three, together with §3’s result that the constitutional triple is indivisible, the doubling is a derivation — given P1–P3 and the constitution, and no more than that.
| Constraints | Algebraic element (grade-class deployed) | Level |
|---|---|---|
| 1 | grade 0 — the scalar | the one constraint, the norm (H₁) |
| 3 | grade 1, spatial — three generators | first created level (H₃) |
| 6 | grade 2, spatial — three (magnetic) bivectors | second created level (H₆) |
| 12 | grade 3, spatial — the spatial pseudoscalar | third created level (H₁₂); Cl(3,0) complete |
| 24 | grade 1, temporal — e₀ enters | the birth of time (H₂₄); Cl(3,1) begins |
| 48 | grade 2, mixed — three electric bivectors (σₖ = eₖe₀) | physical matter, our level (H₄₈) |
| 96 | grade 3, mixed — three mixed trivectors | sub-material (H₉₆) |
| 192 | grade 4 — the pseudoscalar e₀e₁e₂e₃ | Cl(3,1) complete (H₁₉₂) |
The 1 → 3 step does not double — it is constitution. The 3 → 6 → 12 → 24 → 48 → 96 → 192 steps each double — one eigenstate selection per step, each a single binary distinction that doubles the determinate states (§4 opening). The levels are denoted Hₙ, n = the constraint count; each Hₙ is a proper subspace of every Hₘ with m < n.2 The levels deploy the algebra one grade-class per rung, in two halves. The spatial (e₀-free) part goes first, deploying H₃–H₁₂ — generators, then bivectors, then the spatial pseudoscalar — and with that Cl(3,0) is complete. The e₀-involving part follows and mirrors it exactly, being e₀·Cl(3,0) and carrying the same 1, 3, 3, 1 shape. It deploys H₂₄–H₁₉₂: e₀ itself at H₂₄ (§6); the electric bivectors σₖ = eₖe₀, physical three-space, at H₄₈; the mixed trivectors at H₉₆; and the grade-4 pseudoscalar e₀e₁e₂e₃ at H₁₉₂. The descent runs eight levels, H₁ … H₁₉₂, across seven steps — the constitution and six doublings — the last completing Cl(3,1).
Why the physical level is H₄₈. The identification of our level with the σₖ rung runs on a principle §5 states independently: a being’s access is its level’s deployed structure. What deploys at H₄₈ is the relative vectors σₖ = eₖe₀ — and that grade-class is, three ways at once, the algebra of physical embodiment. It is lived space: selecting the timelike direction e₀ splits spacetime into an observer’s time and an observer’s space, whose directions are not the bare eₖ but the relative vectors σₖ — space taken with respect to a worldline, the three-space a body stands in. It is the observable algebra: the σₖ square to +1, anticommute, and act as the Pauli matrices — spin, angular momentum, moment — the whole measurable apparatus of nonrelativistic matter. And it is the electric binding: the σₖ are the electric bivectors, and the electric force is what holds atoms, chemistry, and bodies together. Our observables are exactly this algebra; by §5’s principle, the level whose deployed structure they are is ours. The placement is an identification given two named inputs — the access principle, and the empirical fact that embodied measurement is Pauli/electric through and through — structural content, not a coincidence of counts.
Where it ends. Above the created order stands the eternal apex H₁: nothing is resolved there, because resolution is what the GNST does and the GNST is not yet constituted. Below it the created levels are seven, H₃ through H₁₉₂. The grade structure carrying them is finite. Cl(3,1)‘s highest grade is the grade-4 pseudoscalar, so when that deploys at H₁₉₂ the algebra is spent: no grade 5, and so no further kind to actualize. (The e₀-involving part deploys by grade exactly as the spatial part did.) H₁₉₂ is the maximal decomposition — the whole resolved into its most constrained constituents — and it is where Cl(3,1) gives out. What holds together at that floor, or below it, is past the reach of this algebra: a question of dynamics (A3C), and of what the substrate carries beyond Cl(3,1). The structural result is exact: seven created levels beneath the eternal apex, the kinds exhausted at H₁₉₂.
Time dilation. The temporal constraint here is one of §3’s constitutional triple — the GNST / succession component, one of the three constitutional families the tally partitions into (The structure is fractal, below) — so it is exactly one-third of the total at every created level, and temporal constraints double in lockstep (1, 2, 4, 8, 16, 32, 64). (This is a different count from the generator tally. There e₀ — the aging axis, entering only at H₂₄ — is one axis among four, the 3:1 of §6. The two ratios answer two different questions and do not compete.) A being under more temporal constraints undergoes more collapse events — its succession is finer-grained — so between two levels the local rates stand as their constraint counts: T_A/T_B = n_A/n_B, a ratio of collapse counts that needs no absolute clock beneath it. This is concordance, not derivation: it points the same direction as general relativity’s gravitational dilation (fewer constraints ≈ deeper in the well toward the center ≈ slower local time). The hierarchical effect is nonetheless a discrete step-function distinct from the continuous GR effect — a resonance for A3C to test where it engages the measured universe.
The structure is fractal. The three-fold constitution (temporal mechanism, spatial reference, topological ground) reproduces at every level, the set doubling at each rung as the section’s opening described — three universal constraints joined by three local at H₆, and so on. Each level is the same structure instantiated more specifically; nothing structurally new enters, only specificity increases. Each level contains the full organizational pattern of the levels above, instantiated locally — the holographic character of the hierarchy.
The representation tower — what doubles, exactly. The fractal mechanism has a precise form in representation theory, and it names what is invariant and what grows. The constitutional algebra is simple — Cl(3,0) ≅ M₂(ℂ) (§1) — so it has exactly one irreducible representation up to isomorphism: the spinor module S ≅ ℂ², the three generators acting as the Pauli matrices. Any Hilbert space carrying the algebra’s action therefore factors as
𝒱 ≅ S ⊗ M,
the one irrep S tensored with a multiplicity space M — the algebra acts only on the S leg (as σ ⊗ 1), and M is the multiplicity space — the copies of S. Two multiplicities live here and must be kept apart. The structural one is the cascade’s: how many times the structure is re-instantiated from one level to the next — one at the apex (H₁, undivided), doubling at each fractal step (1 → 2 → 4 → …), a finite tally across the levels. The physical one is different: at the physical level the actual Hilbert space carries unboundedly many copies of the single irrep — the countless instances of matter, one spinor S apiece. The first is the between-level doubling the cascade bookkeeps; the second is the within-level population of a level, and the doubling neither reaches nor needs it. S is identical in every copy — the form fixed the whole way down, only the multiplicity changing: the same structure instantiated more specifically, nothing structurally new entering. The holographic character is the decomposition itself — the whole, S, literally present in every summand of S ⊗ M.
Level and proximity are different coordinates. Neither multiplicity measures a being. The level fixes what a being is constituted by — the algebra acts on the S leg, and S is identical in every copy — so everything at one rung shares its structure, and the rung says nothing about the individual occupying it. What differs between two beings at the same level is not carried by n at all. It is carried by where their states sit on the chain of A0 §2.3: Φ = ⋂ₖ Hₖ is the domain on which every weighted norm is finite at once. A state belongs to some of the Hₖ and not others, so how far along that chain a state has got is a fact about that state. A1 §3 describes a finite being in exactly those terms — not yet fully determinate under every probe, not yet consistent at every scale simultaneously — which is the chain’s criterion said in words.
So the two subscripts count different things and must not be run together: n indexes worlds; k indexes regularity. Two beings at H₄₈ share n and differ in k. A being does not become more actualized by changing constraint level; it becomes more actualized by its state satisfying more of the chain. Development in A1 §3’s sense — the alignment of the H-state with its Φ-core — is motion along k, within a level, and not passage among the levels the cascade counts.
Two conditions bound the picture. First, the clean S ⊗ M picture is the in-algebra descent (H₁–H₁₂, Cl(3,0)); at the crossings the algebra is itself extended — the complement generator e₀ enters (§6), Cl(3,0) → Cl(3,1) — so the physical level carries the irrep at infinite multiplicity inside the time-extended algebra, not the bare Cl(3,0) representation. Second, the multiplicity growth is the fractal mechanism made exact, its doubling grounded in the single binary distinction each collapse makes (§4 opening). But it does not by itself fix the dimension assigned to each Hₙ. That the count runs 3·2ᵏ is the cascade’s own bookkeeping (the table above), and which created instance occupies which slot of M the algebra does not say at all. The structure is defined and the multiplicity is definable; the labelling of the slots — which life, which family, is which copy — is the open question the mathematics hands back to the world.
5. The hierarchy is dimensional, not geographical
The cosmos is commonly pictured spatially — higher realms above, lower below. This is misleading. A being at a level is one constituted by that level’s content, and its reach is that content’s. A higher rung is the richer one — but richer has three senses here, and they do not point the same way, so they are worth separating before the picture is drawn.
Latitude grows going up, and carries no factor. Λ(ψ) is the set of states that may succeed ψ at its rung (A3B), and it contracts monotonically down the descent, so a higher rung admits strictly more successors than a lower: a being at H₂₄ has more room to move than a being at H₄₈. By how much, nothing here says. A3B attaches the cascade’s doubling to the determinate states, not to Λ, and gives Λ the monotonicity alone.
Deployed algebra falls going up, and this is the sense that is exactly countable. The ledger above runs 1, 3, 3, 1, 1, 3, 3, 1 and sums to sixteen, the dimension of Cl(3,1); cumulatively that is 1, 4, 7, 8, 9, 12, 15, 16. So H₂₄ has nine of the algebra’s sixteen elements in force and H₄₈ has twelve. The higher rung is constituted of less algebra, not more — which is the same fact as its having fewer constraints, seen from the other side.
And the state space grows going up, which carries no number. Constraints cut the admissible space down, so the 48-constraint world is a subspace of the 24-constraint space — every physical dimension is present to the H₂₄ level, plus dimensions physical matter cannot reach, the way a plane is a subspace of a 3-D space. That containment is derived; how much larger the containing space is, the cascade does not say, because it does not fix the dimension assigned to each Hₙ (§4).
So the H₂₄ level is not distant; it is freer, constituted of less, and larger — present at the same spatial location.
Perception follows the subspace structure exactly. Higher levels perceive lower (the lower is embedded in the space the higher already occupies); lower levels cannot perceive higher (those dimensions are not in the lower being’s range). Our not perceiving the H₂₄ level is a structural fact and not a limit of instruments: we are constituted by 48-constraint matter, and the H₂₄ level occupies dimensions 48-constraint matter does not sample.
6. The two discontinuities, and the birth of time
The cascade has two qualitative discontinuities, and they differ in exactly one structural respect: whether the entering resource is entailed by what is already deployed. (They are not §2’s two crossings — on §4’s own map those fall between H₃, H₆, and H₁₂, and the H₁₂ → H₂₄ step is not a grade crossing within Cl(3,0) at all; what §2 contributes here is the parity fact that grounds the latency below. The reading of these crossings as kinds of act — and the naming of what supplies the unentailed input — belongs to A4.)
The unentailed crossing, H₁ → H₃. H₁ is eternal in the sense fixed at §3 — prior to the GNST, and so prior to succession. It is A1’s realized nuclear space (Φ, τ_nuclear), and the norm ‖·‖ is its sole constraint: the topology is not a constraint upon H₁ but part of what H₁ is. H₃, with three constraints, is the first created level below it. Nothing inside H₁ generates H₃ — the descent requires a free input — free in the exact sense that no element already deployed entails it, and none of the algebra’s own products yields it. What supplies it is not named here; the naming belongs to A4. It opens the space within which every later step runs, and it is the only step of its kind in the descent.
The entailed crossing, H₁₂ → H₂₄ — where the cascade meets its complement. H₃, H₆, H₁₂ are derivable inside Cl(3,0): H₃ gives the three generators, H₆ the three bivectors, H₁₂ the pseudoscalar I = e₁e₂e₃ — which exhausts the spatial algebra. H₂₄ cannot be reached by any further grade, and the generator it turns on is none of the algebra’s own elements — no product of the three yields a fourth. The crossing is entailed all the same — a further generator entering the products, as at the grade-1 → grade-2 step. But where that step combined a generator already deployed, this one turns on the complement the positive selection casts (below). It is the selection’s own negation, latent while the positive grades deploy (the parity fact, §2) and entering when they exhaust.
The complement, and why e₀² = −1. Any definite selection produces its complement — what is not (P0, the complement posit — the one further posit the descent runs on: it carries e₀’s existence, and, the complement being undifferentiated, its singularity — one direction, not three; the elimination below fixes only its value). On the positive side there is structure: three generators each squaring to +1 — this is Cl(3,0). On the negative side: no differentiation, no structure — a single undifferentiated norm. This is the structural reading of e₀ — the bare negation the selection produces, rather than a temporal principle with an origin of its own. Its value is then fixed by elimination, under the standing given — this exists (the cogito, A1): the fourth generator’s square can only be +1, 0, or −1, and the first two are excluded.
- e₀² = +1 would make it a fourth spatial generator — Cl(4,0), four space dimensions — and Ehrenfest kills it, the same argument that closed the count from above (§1): no stable orbit or atom, no matter that persists, no reality that exists.
- e₀² = 0 would make e₀ null and non-invertible — the inert zero-divisor §1 rejected in the exterior algebra, now in the time slot. It would be a time that cannot act, and so cannot carry the succession of acts the GNST performs through it. Physically it is the Galilean structure — instantaneous, acausal — that A0’s quantum structure, and our own reality, exclude.
That leaves e₀² = −1, and e₀ must enter anticommuting — the complement is orthogonal to the selection by construction, ⟨e₀, eₖ⟩ = 0, and anticommutation is exactly that case of the universal split (§1, The one product, two grades); in the four-generator Cl(3,1), n being even, no grade-1 element can be central in any case.
(Nor does the −1 reopen §1: what §1 ranged over were state-bearing directions, and A0’s metric is positive-definite on those, so under the sign §1 fixed they square to +1 — e₀ bears no states and is not among them. Two roles are in play here and the bullet above turns on the other one: succession is ordered through e₀ — that is what a null time could not carry — while content is borne along the eₖ, and along e₀ there is none. A direction can order the acts without being a direction the acts put content into. So the one −1 lives exactly where §1 leaves room for it.)
So the negation the selection casts is not merely consistent with the Lorentzian −1; the standing given forces it.
This is also why there is one temporal generator, not three — and not zero. A definite selection has a complement: the negation is cast with it, not optional, so the count is not zero. And the complement is undifferentiated, carrying no internal distinction to split it, so it enters as a single direction — one e₀, where the positive side’s differentiation gave three. That singularity is P0’s, declared with it above and not a further inference: Ehrenfest constrains the spatial count and would not by itself forbid a three-dimensional complement, so the posit is what pins the number at one.
The latency, exactly. That e₀ is “latent while the spatial grades deploy” is a closure fact: the spatial Cl(3,0), generated by e₁, e₂, e₃, is closed under the product, so every product of e₀-free elements is again e₀-free and none yields e₀ (which is e₀-involving) — e₀ never appears until the crossing admits it. The even-subalgebra reading says the same through an isomorphic embedding. Cl(3,0) is isomorphic to the even subalgebra of the four-generator Cl(3,1) — grades 0, 2, 4 — generated by the relative vectors σ_k = e_k e₀, for k = 1, 2, 3. Three facts make them a copy of the spatial generators. Each squares to +1: σ_k² = e_k e₀ e_k e₀ = −e_k² e₀² = −(+1)(−1) = +1. They mutually anticommute: σ_jσ_k = e_j e_k = −σ_kσ_j. And their product σ₁σ₂σ₃ = −e₀e₁e₂e₃ is the pseudoscalar, up to sign.
(Where the complexity goes, since §1 selected Cl(3,0) for being complex and the whole Cl(3,1) is not — it is M₄(ℝ), the real algebra of the Majorana form, where the other candidate Cl(1,3) ≅ M₂(ℍ) is quaternionic. The complexity is a property of the even part, and stays exactly where it is needed. Cl⁰(3,1) — grades 0, 2, 4, eight elements — is the copy of Cl(3,0) ≅ M₂(ℂ) just given, and its central imaginary is the grade-4 element σ₁σ₂σ₃, which commutes with everything even and anticommutes with everything odd. That is the requirement met: observables are even — the σₖ deployed at H₄₈, spin and moment and the whole Pauli apparatus — so the phase commutes with everything measurable, which is what makes a global phase unobservable. The odd part carries the dynamics, not the observables; the Dirac operator maps even to odd and is an equation, not a measurement. So §1’s criterion is a test on what A0’s metric writes down directly — the positive-definite spatial part — and its scope ends there.)
But e₀ alone is grade 1 — odd — and even × even = even, so no product within this Cl(3,0) ever yields it: the whole spatial algebra can deploy through H₃–H₁₂ and e₀ as a standalone direction never appears, because it is not reachable from inside the closed even part. It surfaces only once that algebra is exhausted and an odd element is admitted — the even → full extension, which is this crossing. So e₀ stands in two relations at once. It is folded into every spatial generator, since σ_k = e_k e₀ and the σ_k are present from the outset; and it is inaccessible until the grades exhaust, being odd and so outside the even part. These are two faces of one fact, and they give the exact sense of “latent, entering at exhaustion.”
(Two cautions on which even subalgebra is meant. This is not §2’s Cl⁰(3,0) ≅ ℍ, which is grades 0 and 2 of Cl(3,0) itself; here Cl(3,0) sits as the even part of the larger algebra. Nor is it the e₀-free spatial embedding the cascade deploys — it is a different embedding, isomorphic to that one. The latency holds in both.)
What is theorem here is the closure — e₀ is unreachable from within the even part and enters only by admitting the odd, so its appearance is a genuine step — this crossing. As bare algebra, e₀’s status is undetermined — primordial (the spacetime algebra Cl(3,1) fundamental, Cl(3,0) its e₀-free spatial part, e₀ one of four generators from the first) and adjoined (Cl(3,0) fundamental, e₀ the complement admitted here) relate both ways and agree on the closure. But the question is not bare algebra: under the standing given — this exists — the elimination above forces the full algebra to be Cl(3,1), so the framework takes the primordial reading — Cl(3,1) is fundamental, and Cl(3,0), its positive-definite, e₀-free part, is what A0’s metric writes down directly and what the GNST deploys first.
And (3,1) rather than (1,3) follows from §1 and needs nothing further: the sign of the defining relation was settled there, at three directions, by which of the two algebras is a world — so the algebra admitting the complement is Cl(3,1). The two are distinct real algebras rather than two notations — Cl(3,1) ≅ M₄(ℝ) against Cl(1,3) ≅ M₂(ℍ) — and what turns on the difference is the minimal real spinor, ℝ⁴ against ℍ², a distinction that vanishes on complexification and so is invisible to everything metric.
The positive side deploys through its grades and exhausts at H₁₂ — the pseudoscalar completing Cl(3,0). The −1, unstructured, waits — and at the H₁₂ → H₂₄ crossing the two meet: the −1 enters as e₀, and from H₂₄ down the algebra is Cl(3,1). The Lorentzian signature (+,+,+,−) is, on this reading, the algebraic record of the two sides — the positive structure and its complement. From Cl(3,1) the physics follows directly:
- The Dirac equation is spacetime geometric algebra written out — Hestenes’ Spacetime Algebra works it in Cl(1,3), and the Dirac theory also admits a real Cl(3,1) form, the Majorana representation. The framework’s standing given forces the Lorentzian signature — one minus among three pluses (the elimination above) — and the sign §1 settles fixes which of the two algebras it is, so the Majorana form is the one this descent writes.
- E = mc². From Dirac, E² = (pc)² + (mc²)²; at rest, E = mc². What the algebra supplies is the form — the m in it is a parameter the algebra does not fix. The framework’s reading of that parameter: mass is energy at maximum specificity — distributional at H₁, concentrated by eigenstate selection and constraint doubling until, by H₄₈, it is mass in the ordinary sense. And c is the unit the e₀ axis wears against the spatial axes — the conversion the equation’s c² records.
- c is an output, not an axiom. Special relativity stipulates the constancy of c (its second postulate); here c is the coupling that must exist between the positive generators and the complement — its existence and algebraic role derived from the crossing — so SR’s second postulate, that c is the same for every observer, becomes a consequence. Its invariance across observers is what the crossing supplies; whether its value holds fixed across cosmic epochs is a separate question the algebra does not answer — the measurement history and the variable-speed-of-light programs bear on it, and A3C takes it up.
Light is the daughter of the crossing. The electromagnetic field bivectors {e₀e₁, e₀e₂, e₀e₃} (electric) and {e₁e₂, e₂e₃, e₁e₃} (magnetic) span the meeting of the two sides. The photon’s 4-momentum satisfies p² = 0 — the +1 and −1 contributions exactly cancel — so light lies on the boundary between the two sides, the null cone, massless because the balance there is exact. Light is what the meeting of the two sides looks like — the positive structure and its complement in equipoise.
Antimatter. Antimatter is a child of the crossing: a relativistic phenomenon, existing only where e₀ does. Above the crossing the cascade is in Cl(3,0) — no temporal generator, no Lorentzian equation, and so none of the negative-frequency solutions that are antiparticles (a being at H₁₂ or above has no antiparticle counterpart). Below it, the Dirac form admits them of necessity: every actualization in Cl(3,1) carries its charge-conjugate twin as a solution the algebra will not suppress — antimatter the structural shadow of every positive actualization, cast by e₀’s entry.
Definition — succession, and Time
One resolution act = one GNST collapse. The invariant unit — defined, not discovered.
φ succeeds ψ when the vector multiplication between two successive collapses carries ψ to φ. Succession is that carrying: before and after, with no axis to age along.
Time = succession with a worldline — succession borne along the axis e₀ supplies at H₂₄ and below.
The axis is the signature’s; the direction is not. e₀ gives a worldline to be borne along and nothing more; the arrow is the GNST’s, since a collapse resolves potential to actual and does not run back. Succession runs wherever the GNST runs — from H₃ down — so the band H₃–H₁₂ has before and after without aging (*aevum*), and H₁, prior to the GNST, has not succession but is-ness (aeternitas). Which states may succeed a given one is set by the constraints deployed where it stands; that set is A3B’s latitude, and at H₄₈ its bound is the light cone.
The metric birth of time (aging = e₀). Above the crossing the algebra is Cl(3,0): the GNST still runs through the constrained levels H₃–H₁₂, events succeeding one another — “before” and “after” exist — but no direction is the aging axis. This is succession in the strict sense: before and after, with no axis to age along.
Below the crossing e₀ is in the algebra, and the light cone constrains what can influence what. The direction is not the signature’s to give — a Lorentzian metric is symmetric under t → −t, and no arrangement of signs yields an arrow. It is the GNST’s: a collapse resolves potential to actual and does not run back. What e₀ supplies is the axis that succession is then borne along.
This is Time: succession with a worldline. At H₂₄, e₀ has just entered (three spatial generators to one temporal, 3:1) — Lorentzian, aging, but not yet mortal in the H₄₈ sense. At H₄₈ the accessible degrees of freedom (above) have coarsened until they fall within the light cone — locality dominates — and dissolution is borne along a compulsory local worldline. That is aging as physical beings know it, in a body that cannot step off its own decaying path. (What marks the physical level is this closing of the accessible freedom inside the light cone — not a greater share of time, which holds at 3:1 from the crossing down and does not grow.)
The apex H₁ stands above even succession: the GNST is not constituted until H₃, so at H₁ the norm stands alone, with nothing succeeding anything.
7. Conclusion
A0’s space measures; it does not yet act. The operation by which it acts — the product — is forced up to a single named postulate: if actualization reads each state’s is-or-is-not off its square, as a scalar of the metric, it is the Clifford product — the universal property does the rest. The number of directions is the world’s own, and §1 shows it could have been no other, by three exclusions that each hold in both sign conventions. One direction gives a commutative algebra, where conjugation is trivial and nothing acts on anything. Two directions give an algebra whose centre is ℝ — a parity fact, holding for M₂(ℝ) and ℍ alike — so no phase and no complex state. And from four up the cogito’s persisting knower cannot be held, by Ehrenfest’s stability result. Three remains. Only then is the sign answerable, and it is answered by one square: at three the pseudoscalar is central, and whether it squares to −1 or +1 is the difference between a simple complex algebra and a split one. So the sign is settled last, as an output: Cl(3,0), one world, complex, its spinor module A0’s own two-level state.
From there the descent is bookkeeping made exact. §2 separates what a stratum’s own means can reach from what they cannot — parity, the ground of latency. The cascade deploys one grade-class per rung — the constitution of three structures at H₃, then six doublings on the three enumerated posits — seven created levels, H₃ through H₁₉₂, beneath the eternal apex. The complement posit turns on e₀, the elimination fixes its square at −1, and the algebra of the created order is Cl(3,1): the Lorentzian signature, the Dirac form, E = mc², and an invariant speed as consequences of the crossing rather than stipulations before it. The physical level sits at H₄₈ by the σₖ identification — lived space, the observable algebra, the electric binding: one grade-class, three ways.
The inputs are few and named: the actualization postulate, Ehrenfest’s stability, the chain’s cogito, the complement posit, P1–P3, and the two that place H₄₈. Two requirements do work in §1, neither an addition to that list, and they are not owed equally.
That the algebra manufacture a central imaginary is A0’s complex state space asking to be reproduced, and it reaches the algebra by way of the identification i ≡ I, marked as an identification where it is made (§1) and carried as such in the Traceable Account. The exclusion of two directions is doubly grounded — first by the world-premise §1 opens with, this world having three directions and the algebra’s dimension being the world’s; and second, independently of any count, by this identification, since Cl(2,0) is simple and nothing structural disqualifies it. The identification is therefore not what the count hangs on; it is what lets the exclusion stand even where the count is not granted.
That the world be one rather than two halves with no traffic between them is A0’s nuclear domain, a single domain self-consistent under every probe, and it passes through no identification. At three directions both bear on the same fact — the pseudoscalar’s square — so the sign is settled on the second ground and does not depend on the identification.
Everything else is the algebra. The structures stand derived and unnamed — the naming is A4’s, and the measured universe A3C’s.
Traceable Account
Per result: what it stands on (ground), what outside the paper agrees (support), and how it could be wrong (exposure). The head-note Confidence is a reduction of this table; where the two disagree, this table governs.
Math
| Result | Ground | Support | Exposure |
|---|---|---|---|
| The Clifford algebra forced by A0 (§1) | derived given 1 input ¹ | — | beyond this level ⁶ |
| The two crossings (§2) | derived | — | beyond this level |
| Cl(3,0), and the Lorentzian Cl(3,1) (§6) | derived given 3 inputs and one identification ² | — | beyond this level |
| i ≡ I — the algebra’s central imaginary is A0’s (§1) | identification | — | beyond this level |
| The norm ‖·‖ is what gravitates — the SOURCE, not the field (§3) | identification | — | falsifiable — an EM-based direct detection of dark matter (A3C §5) |
| Spacetime = the nuclear topology τ_nuclear (§3) | identification | — | beyond this level |
| The first created level requires exactly three constraints — GNST, the norm, the nuclear topology (§3) | inferred — from what the spectral theorem leaves unfixed | — | beyond this level |
| The cascade’s doubling; the seven created levels H₃–H₁₉₂ beneath the eternal apex H₁ (§4) | derived given P1–P3 + the constitution ³ | — | beyond this level |
| The levels nest — each Hₙ a proper subspace of every Hₘ with m < n (§4) | derived — each collapse selects a subspace | — | beyond this level |
| Local rates stand as constraint counts, T_A/T_B = n_A/n_B (§4) | inferred — the proportionality is posited, not derived; what §6 derives is that ageing is e₀, and that the arrow is the GNST’s | — | open — no cross-level clock |
| Access is the level’s deployed structure; higher levels perceive lower (§5) | posited | — | beyond this level |
| The physical level placed at H₄₈ (§4) | identification given 2 inputs ⁴ | — | falsifiable ⁵ |
| The Dirac equation, and E = mc² (§6) | derived — the form only ⁷ | — | beyond this level |
| c is an output: SR’s second postulate, c invariant across observers, is a consequence of the crossing (§6) | derived — its value across epochs is A3C’s | — | beyond this level |
| Time is succession borne along an axis; the arrow is the GNST’s, not the signature’s (§6) | inferred — the metric is t → −t symmetric ⁸ | — | beyond this level |
| No antiparticle counterpart above the crossing (§6) | inferred — from the Dirac form’s absence in Cl(3,0) | — | beyond this level |
| Mass is energy at maximum specificity — the reading of the m the algebra leaves free (§6) | identification | — | beyond this level |
| The Dirac form, E = mc², and the photon’s p² = 0, as consequences of Cl(3,1) (§6) | derived — the form only ⁷ | — | falsifiable — a departure from the Dirac form (A3C) |
Science — not engaged. The forms are derived at §6; the measured universe is A3C’s.
Theology — not engaged.
Notes. ¹ The actualization postulate. ² Atop the actualization postulate the Clifford selection itself rests on — it gives the Clifford relation rather than the exterior one, and with it the invertibility that distinguishes it from the exterior algebra (§1) — three: Ehrenfest stability; the chain’s standing from A1’s cogito (existence, with persistence following inline, §1); and the complement posit P0 (§6) — a selection casts its complement, carrying both e₀’s existence and, the complement being undifferentiated, its singularity, while the elimination fixes only its value. And one identification, named because it does work the structure alone does not: the algebra must manufacture a central imaginary to reproduce A0’s complex state space, and it reaches the algebra by the identification i ≡ I (§1, The real and the complex). It gives the second, count-independent exclusion of two directions — Cl(2,0) being simple, nothing structural disqualifies it — but the count does not depend on it: two directions are excluded first by the world-premise of §1, this world having three. The sign selection at three directions does not pass through that identification, splitness giving it an independent ground. ³ P1–P3, §4. ⁴ §5’s access principle; and that embodied observables are the Pauli/electric σₖ algebra. ⁵ A measured dark-to-baryonic ratio inconsistent with five levels standing above ours — H₂₄, H₁₂, H₆, H₃ and the apex H₁, which is counted because the norm does not stop at the created order (A3C §5). The ratio measures which rung we occupy: 4:1 would place us at H₂₄, 6:1 at H₉₆. ⁶ “Beyond this level” means refuted by error, not by observation. ⁷ Identifications and readings marked where made (§3, §6). ⁸ The signature supplies the light cone and the axis; the directedness is the GNST’s, a collapse resolving potential to actual.
References
-
Companion paper. Modeling Reality as a Gelfand Triple (A0). [The complex inner-product space, the positive-definite metric absolute, the associative nuclear domain, and the actualization operation (the GNST) on which this paper’s algebra is built.]
-
Lang, S. (2002). Algebra (3rd ed., Graduate Texts in Mathematics 211). Springer. [The definition of an algebra over a field — a vector space carrying a bilinear multiplication. See also Bourbaki, Algebra I (Ch. III), and Dummit & Foote, Abstract Algebra (3rd ed.).]
-
Chevalley, C. (1954). The Algebraic Theory of Spinors. Columbia University Press. [The foundational algebraic theory of Clifford algebras and their spinor representations.]
-
Lawson, H.B., and Michelsohn, M.-L. (1989). Spin Geometry. Princeton University Press; Lounesto, P. (2001). Clifford Algebras and Spinors, 2nd ed. Cambridge University Press. [The classification of the Clifford algebras (Bott periodicity): in this paper’s convention the single, complex, positive-definite members are Cl(n,0) with n ≡ 3 (mod 4), Cl(3,0) ≅ M₂(ℂ) acts on the spinor module ℂ², and Cl(0,3) ≅ ℍ ⊕ ℍ is split with centre ℝ ⊕ ℝ. ⚠ Lawson & Michelsohn write the defining relation with the opposite sign (Preliminaries), so their printed labels are the transpose of these — their Cl(n,0) is this paper’s Cl(0,n). The classification is the same theorem under either labelling; Lounesto states it in the convention used here.]
-
Hestenes, D., and Sobczyk, G. (1984). Clifford Algebra to Geometric Calculus. D. Reidel. [The geometric product read at its grades, ab = ⟨a,b⟩ + a∧b, and the geometric-algebra idiom used throughout; for the Cl(3,1)/Dirac form of §6, Hestenes, Space-Time Algebra (Gordon & Breach, 1966).]
-
Lounesto, P. (2001). Clifford Algebras and Spinors (2nd ed.). Cambridge University Press. [The grade structure, the pseudoscalar, and the even subalgebra Cl⁰(3,0) ≅ ℍ used in §2’s asymmetry (§2).]
-
Ehrenfest, P. (1917). “In what way does it become manifest in the fundamental laws of physics that space has three dimensions?” Proceedings of the Amsterdam Academy 20, 200. [The original stability argument: in four or more spatial dimensions stable planetary orbits and stable atoms fail — the result that closes §1’s count from above. The critical coupling at four dimensions is (d−2)²/4, equal to 1 there — the threshold §1 turns on.]
-
Tangherlini, F.R. (1963). “Schwarzschild field in n dimensions and the dimensionality of space problem.” Il Nuovo Cimento 27, 636. [The generalization to n dimensions: in four or more spatial dimensions no stable bound orbit exists; the potential admits no stable minimum.]
-
Barrow, J.D., and Tipler, F.J. (1986). The Anthropic Cosmological Principle. Oxford University Press. [The classical- and quantum-stability dimensionality arguments collected: bound states are stable only at three spatial dimensions.]
-
Salthe, S.N. (1985). Evolving Hierarchical Systems: Their Structure and Representation. Columbia University Press. [Hierarchy theory’s elaboration of nested levels of nature — the descent shape of §4, developed qualitatively and without a level count.]
-
Michel, L. (1980). “Symmetry defects and broken symmetry. Configurations. Hidden symmetry.” Reviews of Modern Physics 52(3), 617–651. [Symmetry breaking as motion in the lattice of isotropy subgroups: the standard statement that a breaking hierarchy is a chain of nested subgroups.]
-
Sala, P., Rakovszky, T., Verresen, R., Knap, M., and Pollmann, F. (2020). “Ergodicity Breaking Arising from Hilbert Space Fragmentation in Dipole-Conserving Hamiltonians.” Physical Review X 10, 011047. arXiv:1904.04266. [Local constraints fracturing a Hilbert space into dynamically disconnected subsectors — the nearest existing treatment of this paper’s object, and lateral where §4’s is a descending nest.]
-
Klebanov, I.R., and Strassler, M.J. (2000). “Supergravity and a confining gauge theory: duality cascades and χSB-resolution of naked singularities.” JHEP 08 (2000) 052. arXiv:hep-th/0007191. [A cascade whose step (N → N−M) is derived and which terminates in the infrared — the closest precedent for a derived, stopping descent; additive rather than multiplicative, and one-sided.]
Papers ← Paper A2 · Braided ← Chapter A3A
Footnotes
-
A descending hierarchy of nested levels is not a new shape, and the claim here is not that it is. Hierarchy theory has elaborated it at length — Salthe (1985) sets out a hierarchy of nature whose levels nest by degree of specification — and in physics, symmetry breaking is standardly described as motion down a chain of nested subgroups (Michel, 1980). Closer to this paper’s object, the Hilbert-space fragmentation literature has constraints fracturing a Hilbert space directly (Sala et al., 2020) — though laterally: a symmetry sector shatters into exponentially many disconnected subsectors, siblings on one level rather than a chain of proper subspaces, and the count grows with system size rather than terminating. Closest in mechanism is the duality cascade of Klebanov and Strassler (2000), in which each step is derived from the one before and the chain stops — but the step is additive and the cascade is one-sided, infinite in the ultraviolet. What none of these supplies is what §4 claims: a factor between levels derived from stated posits, and a depth fixed by the exhaustion of the structure that generates the levels. ↩
-
The Framework’s Hₙ subscripts denote the Hilbert space indexed by the number of constraints operating at each level; the subscript numbers — H₃, H₆, H₁₂, H₂₄, H₄₈, H₉₆, H₁₉₂ — appear in order with a doubling between each step. These subscripts come from the cascade’s structure, not from observation: the constraint doubling is carried by information — each collapse a single binary distinction, which doubles the determinate states, the constraints doubling with them because the constitutional triple is indivisible (§3) — a derivation resting on P1–P3 and that constitution. The notation describes only between-level structure; it does not attempt to name within-level refinements. ↩