Naialu Institute of Motion Dynamics · Formal Series
NMC-MATH-001
Algebra of System Operations
Epistemic labels used throughout
Notation and Carrier Sets
Everything below is stated over one ground object, the particle stream, and one derived map, the metric signature. The two operators act on that ground object. The three equivalence relations decide when two results count as the same.
A particle stream is a finite nonempty sequence of positive integers, produced by running a legal name and a Spiral Calendar date of birth through the cipher (Engine Spec, Layer 0). Write a stream as σ = (p₁, p₂, …, pₙ), length |σ| = n.
A system S is a stream together with its derived metric signature. Write σ(S) for the stream of S. Concatenation of streams is written σ ⊕ τ and is the ordinary juxtaposition of sequences.
Every particle is a positive integer, so PT ≥ 1 for every system and the digital root below is always well defined.
The primitives, verbatim from the locked Engine Spec, Layer 1:
Two families are immediate from these formulas and matter for everything downstream:
- PT-derived, stream-order independent: PT WC FS Π T C AC A FSat. Each is a function of
PTandWC = n−1only. - Δ-derived, stream-order dependent: Δ τ v M XEV. Each carries
Δ, which reads the shape of the sequence, not its sum. XEV inherits order dependence throughτ.
For systems X and Y:
These nest by refinement. Identical streams have identical signatures, and FS is one entry of the signature, so
Strict equality is the default for every algebraic claim. The two weaker relations are named so that a property that fails strictly can be re-asked at a coarser layer. This layering is the reason the identity question has three different answers rather than one.
Note on ordering. Metric-equivalence refines FS-equivalence, so the correct chain is strict then metric then FS. The results below are organized along this chain, and the payoff is that structure appears only as the relation coarsens.
Object Domain
Let S be the set of all systems. The domain carries four objects.
- System (element of
S): one stream and its signature. - Relationship (element of
R): the typed structure returned by interaction on an ordered pair. Defined by the SPEC-001 v0.2 pipeline and imported here as a codomain, not re-derived. - Interaction graph: the two-node typed multigraph assembled by that pipeline, generalizing to an N-node population network.
- Fused system: an element of
Sreturned by fusion. Its internal structure is settled by the next definition.
A fused system is flat. It is an ordinary element of S: a single stream with no privileged internal boundaries. It does not remember that it was assembled from blocks.
Formally, fusion produces the stream σ(A) ⊕ σ(B) and reads it as one system. The block history [A.date | A.name | B.date | B.name] may be recorded as provenance metadata for audit, but it is not part of the mathematical object and no operator may read it.
This is what makes re-fusion well defined: because a fused system is an ordinary system with an ordinary stream, it can serve as an operand again, and Φ(Φ(A,B),C) is a legitimate expression.
Operators
The framework carries two operators over one object domain, which is what makes it a formal system rather than a set of calculations. One preserves structure and reports how two systems relate. One transforms structure and reports what single body they become.
Interaction takes an ordered pair of systems and returns a relationship: the typed relational contributions, mechanisms, and emergent interaction properties of the SPEC-001 v0.2 pipeline. It leaves the operands intact. The two systems remain two systems. I asks how two relate.
Its internal construction is specified elsewhere and is not an object of proof here. In this document I appears only as the second operator whose interaction with Φ is the flagship open problem.
Fusion takes an ordered pair and returns one flat system whose stream is the ordered concatenation of the two operand streams. Φ asks what single body two systems become.
The metric signature of Φ(A,B) is read off σ(A) ⊕ σ(B) by Def 0.2, exactly as for any system.
Convention on the operand order
The algebra is developed for Φ as left-first concatenation. The framework's oldest-first rule is an external normalization that selects which operand is written first, keyed to date of birth. By Theorem 3.4 it changes nothing on the PT-derived layer and pins only Δ. Its status for iterated fusion is Conjecture 4.3, not a result.
Proven Properties
Six results, each proven analytically from Def 2.2. The first is the engine that drives the rest: fusion is addition on the stream, so any metric that is a function of the sum inherits the algebra of addition, and any metric that reads the shape does not.
PT and WC are additive under fusion, and FS is a homomorphism onto the cyclic group of order nine.
Proof
Concatenation places every particle of A and every particle of B into one stream, so PT(Φ(A,B)) = PT(A) + PT(B). Thus PT : (streams, ⊕) → (ℤ≥₁, +) is a monoid homomorphism.
Length adds, so |σ(A) ⊕ σ(B)| = |σ(A)| + |σ(B)|, giving WC(Φ(A,B)) = WC(A) + WC(B) + 1. WC is affine-additive, order independent.
The digital root satisfies dr(a + b) = dr(dr(a) + dr(b)), which is casting out nines: dr is the quotient map (ℤ≥₁, +) ↠ (ℤ/9ℤ, +) under the identification of the residue class of 0 with the label 9. Therefore
is a monoid homomorphism onto a group, with FS(Φ(A,B)) = dr(PT(A) + PT(B)). □
Fusion strictly grows the system. For all A, B ∈ S:
Equivalently, adjoining any operand E ∈ S strictly increases both counts: WC(Φ(A,E)) > WC(A) and PT(Φ(A,E)) > PT(A).
Proof
By Theorem 3.1, WC and PT are additive under fusion. Every system is a nonempty stream of positive integers (Def 0.1), so WC(A), WC(B) ≥ 0 and, with the one junction pair created by concatenation, WC(Φ(A,B)) = WC(A) + WC(B) + 1. Likewise each operand carries PT ≥ 1, so PT(Φ(A,B)) = PT(A) + PT(B) ≥ PT(A) + 1. Both totals strictly exceed either operand alone. □
Role
This is the single monotonicity fact consumed by every non-existence result below. Theorem 3.7 (no identity), Corollary 3.7.1 (no inverses), and the conical claim in Corollary 3.8.1 all appeal to it rather than re-deriving that fusion lengthens.
Φ(A,B) ∈ S for all A, B ∈ S.
Proof
The concatenation of two finite nonempty positive-integer sequences is a finite nonempty positive-integer sequence, hence a valid stream (Def 0.1). By Def 1.2 it is read flat as an ordinary system. □
Φ(Φ(A,B),C) = Φ(A,Φ(B,C)) under strict equality, and therefore under metric-equivalence and FS-equivalence as well.
Proof
By Def 1.2 the fused system is flat, so σ(Φ(A,B)) = σ(A) ⊕ σ(B) with no residual boundary. Applying Def 2.2 twice:
Sequence concatenation is associative, so both equal the single sequence σ(A) ⊕ σ(B) ⊕ σ(C). Every metric in Def 0.2 is a function of the resulting sequence, so identical sequences force identical signatures. The two sides are strictly equal. □
Remark, why this reaches even Δ
Associativity holds for the order-dependent metrics too, because re-parenthesizing does not reorder the stream. The final sequence is literally the same, so its reversal count Δ is the same. Order sensitivity is a fact about commuting operands, not about associating them. This is the structural difference between the next two results.
Fusion commutes on the PT-derived layer and fails to commute on the Δ-derived layer. Precisely, for all A, B:
Φ(A,B)andΦ(B,A)agree on every PT-derived metric: PTWCFSΠTCACAFSat. In particular the fused archetype is permutation invariant, soΦcommutes up to FS-equivalence.Φ(A,B)andΦ(B,A)need not agree on ΔτvMXEV. SoΦdoes not commute up to metric-equivalence, and does not commute strictly.
Proof of the invariant part
PT is a sum and addition is commutative, so PT(Φ(A,B)) = PT(A)+PT(B) = PT(Φ(B,A)). Length is order independent, so WC agrees. Every PT-derived metric is a function of PT and WC alone (Def 0.2), so all of them agree, including FS = dr(PT). □
Proof of the variant part
It suffices to exhibit one pair on which Δ differs. This is done in Counterexample 3.5, which also computes the gap exactly by Proposition 3.6. A single witness retires the universal claim of commutativity for the Δ-derived layer. □
Reading
The archetype of a fused body does not depend on which partner is written first. Only the abruptness terms do. Oldest-first is therefore a convention that pins Δ and touches nothing on the archetype layer.
Concatenation creates exactly one new adjacency, the junction pair (last σ(A), first σ(B)). Writing sᵀ for the sign of the final step of A, sᵇ for the sign of the first step of B, and s* for the sign of the junction step first σ(B) − last σ(A), then whenever none of these three steps is flat:
where [·] is 1 when true and 0 when false.
Proof
Reversals are transitions between consecutive step signs. All internal steps of A and of B survive concatenation, contributing Δ(A) + Δ(B). The only new material is the junction step s*, inserted between A's last step and B's first step. It creates precisely two new consecutive-step transitions, (sᵀ, s*) and (s*, sᵇ), each a reversal exactly when its two signs differ. Summing gives the formula. □
The asymmetry of the two bracket terms under swapping A and B is the entire source of non-commutativity in Δ. The full formula including flat steps and run merging is Conjecture 4.2.
Take σ(A) = (1,3,2) and σ(B) = (4,5). Internal reversals: Δ(A)=1 (rise then fall), Δ(B)=0.
By Proposition 3.6, junction A∣B: sᵀ=−, s*=sign(4−2)=+, sᵇ=+, giving 1 + 0 + [−≠+] + [+≠+] = 1+0+1+0 = 2. Junction B∣A: sᵀ=+, s*=sign(1−5)=−, sᵇ=+, giving 0 + 1 + [+≠−] + [−≠+] = 0+1+1+1 = 3. Both match the direct count.
PT = 15 and WC = 4 for both orderings, so FS = dr(15) = 6 and every PT-derived metric agrees, confirming Theorem 3.4. Only τ = PT·Δ splits: 30 versus 45. A single asymmetric pair suffices. □
There is no system E with Φ(A,E) = A for all A under strict equality, and none under metric-equivalence.
Proof
Both cases are immediate from Length Monotonicity (Lemma 3.1.1).
Strict: the lemma gives WC(Φ(A,E)) > WC(A) for every operand E, so the two signatures already differ in their WC entry and cannot be strictly equal.
Metric-equivalence: the lemma gives PT(Φ(A,E)) > PT(A), so the signatures differ in their PT entry and are not metric-equivalent.
An identity would have to leave both PT and WC fixed. The lemma forbids that for every operand, so no identity exists at either layer. □
Corollary 3.7.1
No inverses exist under strict equality or metric-equivalence. By Length Monotonicity (Lemma 3.1.1) the fusion monoid is conical, so no element other than an identity could be invertible, and by Theorem 3.7 there is no identity.
Under FS-equivalence, the archetype coordinate of fusion is exactly (ℤ/9ℤ, +), with FS9 as the label of the residue class 0. Concretely:
- Identity. The class of systems with
FS = 9satisfiesFS(Φ(A,E)) = FS(A)for allA, on both sides. FS9 is the two-sided FS-identity, and it is the only one. - Inverses. Every FS-class has an FS-inverse: the inverse of FS
ais FS9−afora ∈ {1,…,8}, and FS9 is self-inverse. The inverse pairs are1↔8, 2↔7, 3↔6, 4↔5, 9↔9, the complement-to-nine pairs. - Generator. FS1 generates the whole group. The generators are the classes coprime to 9, namely FS
{1,2,4,5,7,8}. FS3 and FS6 generate only the order-three subgroup{FS3, FS6, FS9}. - Idempotents.
FS(Φ(A,A)) = FS(A)holds only for FS9. It is the unique FS-idempotent.
Proof
By Theorem 3.1, FS(Φ(A,E)) = dr(PT(A) + PT(E)) = dr(dr(PT(A)) + dr(PT(E))) = dr(FS(A) + FS(E)). So on the FS coordinate, fusion is addition of digital roots, which is addition in ℤ/9ℤ with the residue 0 written as 9.
Identity. dr(FS(A) + 9) = dr(FS(A)) because adding a multiple of nine does not change a digital root, so any E with FS(E)=9 fixes FS(A). Commutativity of addition gives the same on the left. If FS(E) = k ≠ 9, choose A with FS(A)=9; then FS(Φ(A,E)) = dr(9+k) = k ≠ 9 = FS(A), so no other class is an identity.
Inverses. In ℤ/9ℤ the inverse of a is −a ≡ 9−a, and 0 is self-inverse. Translating labels gives the pairs stated.
Generator. An element generates ℤ/9ℤ iff it is coprime to 9, giving {1,2,4,5,7,8}. gcd(1,9)=1, so FS1 generates.
Idempotents. 2a ≡ a (mod 9) iff a ≡ 0, that is FS9 only. □
Corollary 3.8.1, the structure does not lift
The group lives entirely in the homomorphic image. It is not a subalgebra of S. Downstairs, (S, Φ) is a conical (Lemma 3.1.1), associative, non-commutative concatenation monoid with no identity and no inverses (Thm 3.3, 3.4, 3.7). The FS map carries it onto a finite abelian group, and every group feature above is a fact about that quotient, not about any pair of actual bodies.
Remark 3.8.2, semantic resonance, held separate
That FS9 Crystallization is the fusion identity on archetype is a theorem of modular arithmetic, not an interpretive claim. Its resonance with the framework reading of FS9 as the phase gate that re-seeds is an Observation, recorded here and nowhere used as a premise. The kernel of the FS map is exactly the FS9 class, the streams with PT ≡ 0 (mod 9).
Remark 3.8.3, the group does not respect the geometry partition
The complement-to-nine inverse pairing crosses the direction classes. The pair 3↔6 couples a spiral archetype with an OUT archetype, and 2↔7 sits inside IN while 1↔8 sits inside OUT. The cyclic group structure and the OUT / IN / spiral / none partition are independent overlays on the same nine labels. Neither refines the other.
Summary exhibit: property by equivalence layer
| Property of Φ | Strict | Metric-equiv | FS-equiv |
|---|---|---|---|
| Closure | holds | holds | holds |
| Associativity | holds | holds | holds |
| Commutativity | fails (Δ) | fails (Δ) | holds |
| Identity | none | none | FS9 class |
| Inverses | none | none | 9−a |
| Idempotents | none | none | FS9 only |
| Structure | conical monoid | conical monoid | ℤ/9ℤ |
Reading the table down each column: nothing invertible appears until the relation is coarse enough to forget length. Reading across the identity and inverse rows: structure is not present in the systems and then discovered, it is created by the quotient. This is the concrete payoff of the three-tier equivalence hierarchy.
Open Properties
Stated precisely and left open. None of these is answered here, and none is answered from data.
As sequences, (streams, ⊕) is the free monoid on the particle alphabet, so all identification happens in the metric map. Characterize the fibers of the full signature map: when are two distinct streams metric-equivalent, and what is the structure of those classes under fusion. The non-freeness of the system algebra lives entirely here.
Proposition 3.6 covers the strict-sign case. Extend it to flat steps at the junction and to run merging, where a flat run can bridge two same-sign or opposite-sign runs. Deliver a closed form for Δ(Φ(A,B)) valid for all streams, then read off the exact non-commutativity gap Δ(Φ(A,B)) − Δ(Φ(B,A)) in closed form.
Binary fusion is left-first. The oldest-first normalization keys the operand order to date of birth, but a fused body is flat and carries no canonical date, so oldest-first does not extend to iterated binary fusion. The natural canonical n-ary object is sort-then-concatenate on the multiset of persons, which is order independent by construction. Decide whether sort-based n-ary fusion is adopted as the canonical multi-body operator, and prove it agrees with binary fusion exactly on the PT-derived layer while pinning Δ deterministically.
Does I(Φ(A,B), C) relate in any stable way to the original coupling I(A,B). This is the one question that mixes both operators over the domain. It is the reason the framework is an algebra of two operations and not one, and it is deferred to a document that formalizes I as a closed map before the relation can be posed as a proposition.
Theorem 3.8 gives the kernel of the FS map as the FS9 class. Whether membership in that kernel, or the group-theoretic distance between two systems on the FS coordinate, carries any predictive load is an empirical question. It is named here and answered nowhere in this document.
Empirical Algebra Pass
Reserved. Empty by design.
No dataset results are recorded here. An empirical pass, for example the observation that across a tested population fusion appears associative with respect to FS but not with respect to Δ, belongs in a separate later document written in evidence language. Observation is not dressed as proof. Data can falsify a proposition through a single counterexample, or stand as evidence for a conjecture, but it can never establish a theorem in this document.
