Motives, Periods, and Amplitudes: The Coalgebraic Anatomy of Physical Representation
Abstract
This is Part II of the modular series A MathPhysics Representation Library, whose governing thesis is that a large class of physical quantities are not merely described by mathematics but are realized from structured mathematical information along an explicit pipeline Part I built the ambient representation stack and left the realization channels and the decomposition operation as abstract data. This paper supplies their first fully concrete instantiation. We formalize motives as the functional essence of a mathematical object (a status- semantic layer, carefully distinguished from Grothendieck’s standard definition), we make the realization channels the named cohomological functors for , and we make the observable the period map . The central object is the motivic amplitude object with , equipped with a coaction and a cobracket . We prove: (T1) a Conditional Amplitude Decomposition theorem stating that a realization channel, realized as a homomorphism of the motivic-Galois Hopf algebras, transports the coaction into a decomposition of the realized amplitude; (T2) that the weight-graded primitives cogenerate the Goncharov Lie coalgebra , with weight-one primitives identified with ; (T3) multiplicativity of the period pairing under direct sum and tensor product; (T4) a Non-Tate Obstruction limitation theorem showing that elliptic and other non-Tate motivic lifts escape the polylogarithmic anatomy; and (T5) a conditional (status /) compatibility statement for the cosmic Galois action. Throughout we retain the S/H/P epistemic-status discipline of Part I, and we exhibit the coalgebraic anatomy as the concrete arithmetic mechanism behind cuts, discontinuities, factorization, and residues. We accompany the theory with a compiling Haskell formalization of the coalgebra/coproduct structure and an idiomatic Lean sketch of the period/amplitude/coalgebra types.
1 Introduction
1.1 The representation library and where Part II sits
The series A MathPhysics Representation Library formalizes the idea that mathematics functions as the syntax of physical representation: physical observables are obtained as realizations of structured mathematical information. The program is deliberately modular. Rather than positing a single universal functor from all of mathematics to all of physics — which we argue in 11 is unlikely to be well defined — we build a ladder of domain-indexed modules, each of which is defensible on its own and each of which composes with its neighbours to add an emergent structural property.
Part I introduced the ambient formalism: a category of mathematical domains, a category or higher category of structures in each domain, a category of physical-representation targets, and the bracket notation for the physical representation assigned to . Its central construct is the realization pipeline
This paper provides that content in the motivic channel. We take to be “pass to the motive” , we take to be the classical realization functors of a motive (Betti, de Rham, Hodge, étale), and we take to be the period map . Under this dictionary the compressed master slogan of the series, , becomes the motivic refinement
1.2 The emergent property added by Part II
In the language of the modular ladder, each level adds an emergent property that its predecessor could only gesture at. Part I could speak of “realization channels ” only in the abstract. Part II makes those channels concrete and numerically checkable: Betti/de Rham/Hodge/étale realizations are honest functors with honest comparison isomorphisms, and periods are honest numbers one can compute. More importantly, Part II endows observables with an internal coalgebraic anatomy: the coaction and cobracket decompose an amplitude into primitive informational pieces. Where Part I’s Decomposition Axiom merely asserted that “compositional observables admit a decomposition operation,” Part II identifies that operation with the Goncharov coproduct and its cobracket, and proves structural theorems about it. This is the sense in which coalgebras are the decomposition laws of physics.
1.3 The epistemic-status discipline
Following Part I we tag every dictionary entry and every claim with one of three epistemic-status labels, and we regard these labels as part of the formalism, not as decoration.
| Label | Meaning | Canonical example |
|---|---|---|
| standard use in mathematics or mathematical physics | Goncharov coproduct on polylogarithms | |
| strong heuristic representation-dictionary entry | motive as “functional essence” | |
| speculative philosophical ontology | physical reality as motivic realization |
Composite labels such as / and / occur where an entry is standard as pure mathematics but heuristic or speculative in its physical reading. We preserve these composite labels verbatim. The discipline forces us to state, for every amplitude class, whether the motivic coaction is an established fact (status , e.g. Panzer–Schnetz on periods [10]) or a conjectural extension (status /, e.g. the cosmic Galois group [8]).
1.4 Contributions
A precise formulation of the functional-essence reading of a motive as a status- semantic layer over Grothendieck’s standard notion, with a boxed caution (10) that prevents the reading from being mistaken for the mathematical definition.
An instantiation of Part I’s realization pipeline (1) in the motivic channel, identifying with the motive, the cohomological realizations, and the period map (3, 4).
A self-contained proof of the Conditional Amplitude Decomposition theorem (22), stating that a realization channel — realized as a homomorphism of the motivic-Galois Hopf algebras induced by a Tannakian realization functor — transports the coaction on to a decomposition of the realized amplitude, together with an explicit statement of where the S/H/P labels bite.
A structure theorem for the Goncharov Lie coalgebra (26): the weight-graded primitives cogenerate , and weight-one primitives are .
Multiplicativity of the period pairing (14) and a Non-Tate Obstruction limitation theorem (31) giving a citable, rigorous form of the series’ Limitation 4.
A compiling Haskell formalization of period data, motivic amplitude objects, coactions and cobrackets, and the Goncharov Lie coalgebra, together with an idiomatic Lean type-signature sketch (10).
1.5 Outline
2 recalls the Part I framework and fixes motivic notation. 3 treats motives as functional essence and the realization functors. 4 develops periods and proves multiplicativity of the period pairing. 5 introduces motivic amplitude objects, coalgebraic anatomy, and proves the Conditional Amplitude Decomposition theorem. 6 constructs the Goncharov Lie coalgebra and proves the generation theorem. 7 gives the physics dictionary — coproducts, cuts, factorization, residues, positive geometries, and the Connes–Kreimer Hopf algebra. 8 proves the Non-Tate Obstruction theorem and states the conditional cosmic Galois compatibility. 9 collects the results. 10 presents the Haskell and Lean formalizations. 11 discusses limitations and the bridges to Parts I and III, and 12 concludes.
2 Mathematical Framework
2.1 Recollection of the Part I formalism
We briefly recall the ambient objects; see Part I for details.
Definition 1 (Representation entry, Part I). A representation entry is a quadruple where is a mathematical structure, is a proposed physical representation, is a semantic translation datum (a functor, natural transformation, equivalence class of models, interpretation map, or weaker relation), and is the epistemic-status label.
Definition 2 (Representation prestack / stack, Part I). A representation prestack is a pseudofunctor assigning to each domain a groupoid of representation entries and to a refinement a restriction functor . Given a Grothendieck topology on , is a representation stack if compatible families of local entries glue to global entries uniquely up to coherent isomorphism.
Part I imposes four axioms on the pipeline (1), each of which Part II will instantiate concretely:
Axiom 3 (Realization). .
Axiom 4 (Equivalence). Mathematical descriptions related by an equivalence in the automorphism/gauge groupoid of a representation entry determine the same physical content at the chosen observational level.
Axiom 5 (Locality/descent). Physical representations must be locally assignable and globally coherent; failure of descent is an obstruction, anomaly, or missing boundary datum.
Axiom 6 (Decomposition). Compositional, period-like observables admit a decomposition operation (boundary, coproduct, coaction, spectral sequence, factorization) revealing lower-level informational pieces.
The whole of Part II may be read as the statement: in the motivic channel, 6 is the Goncharov coproduct, and 3 is the period map.
2.2 Motivic notation and standing conventions
We work over a base field of characteristic zero, usually a number field, with a fixed embedding . We write for smooth quasi-projective varieties over . We assume a Tannakian category of mixed motives (or, in the concrete cases we use, the category of mixed Tate motives, which is unconditionally available over number fields by Levine, Deligne–Goncharov, and Voevodsky, and whose period side is controlled by Brown [6]). All tensor products without a subscript are over . For a graded vector space we write for and call the weight.
Definition 7 (Realization functor). A realization is an exact -functor to a Tannakian target . The four classical realizations are the Betti realization (singular cohomology, target -vector spaces with a Hodge/weight filtration), the de Rham realization (algebraic de Rham cohomology, target filtered -vector spaces), the Hodge realization (mixed Hodge structures), and the étale realization (-adic Galois representations).
The Betti and de Rham realizations of a motive are related by the comparison isomorphism
Remark 8 (The motivic channel of the pipeline). 7 makes precise the sense in which Part II instantiates Part I’s : the abstract index of (1) now ranges over the honest set , and each is a genuine functor rather than a placeholder. The remaining “operational/quantum/thermodynamic” channels of Part I are treated in later parts of the series.
3 Motives as Functional Essence
3.1 The standard notion and the heuristic layer
Grothendieck’s motives are the universal cohomological invariants of algebraic varieties: is designed so that every “reasonable” (Weil) cohomology theory factors through a single realization functor out of the category of motives. Betti, de Rham, -adic étale, and crystalline cohomologies are then shadows of one motivic object.
Our series overlays on this a semantic reading, which we flag as heuristic.
Definition 9 (Functional-essence interpretation, status ). The functional-essence interpretation assigns to a variety the semantic datum read as “the invariant, pre-numerical structure from which the various functional readings, cohomological realizations, periods, and numerical shadows of are realized.” In the pipeline (1) this is the concrete choice .
Remark 10 (Caution: a motive is not literally “the functions of” ).
It is essential not to misread 9. Mathematically, a motive is not “the functions of” an object, and it is not the ring of functions or any of its variants. A motive is a universal object unifying the different cohomology theories of ; it is the source of ’s cohomological realizations, not its algebra of functions. The reading “functional essence” is a proposed semantic layer of this project (status ), introduced to align motives with the pipeline’s slogan . Every use of the phrase “functional essence” below inherits the status- label of 9 and must not be taken as a mathematical identity.
3.2 The realization chain
Under 9, Part I’s pipeline specializes to the motivic realization chain
- →
- ↓
- →
- ↓
- →
- ↓
- =
- →
- →
- →
Example 11 (The logarithm as a weight-two period). Let and consider the relative cohomology motive for . Its de Rham side is spanned by the class of and its Betti side by the class of the path from to . The period is This is the motivic incarnation of the statement “ is a period”; the motive is an extension of by , i.e. the Kummer motive. In the functional-essence reading (status ), is a numerical shadow of the Kummer motive, and the Kummer motive is the pre-numerical source common to all the ways one might present .
Example 12 (Multiple polylogarithms). The multiple polylogarithms are periods of the motivic fundamental group of (and its cyclotomic variants). Their weight is and their depth is . Multiple zeta values (with ) are the corresponding special values. These are the canonical periods populating amplitude computations, and their motivic lifts generate the mixed Tate part of the coalgebra we build in 6.
4 Periods and the Period Pairing
4.1 Period data and the period map
Definition 13 (Period datum). A period datum is a quadruple where is a smooth algebraic variety over , is a normal-crossings divisor (the boundary), is an algebraic differential form on , and is a relative cycle in of complementary dimension. Its period is
Period data are the concrete carriers of the pairing between de Rham data () and Betti data (); the number is exactly a matrix entry of the comparison isomorphism (3) for the motive . The Kontsevich–Zagier period ring is the -algebra generated by all such numbers [1]; it contains , , , , and the amplitude constants of perturbative quantum field theory.
4.2 Multiplicativity of the period pairing
The following theorem is the rigorous backbone that licenses treating as respecting the physical amplitude’s additive and multiplicative structure.
Theorem 14 (Multiplicativity of the period pairing). Let and be period data.
(Additivity.) If and have the same and disjoint cycles , then the disjoint-union datum satisfies . More generally the period map is -bilinear in .
(Multiplicativity.) Define the product period datum Then
Remark 15 (What is and is not being asserted). The mathematical content of 14 is the pair of identities (i) and (ii) — in particular the multiplicativity (ii), which is the analytic shadow of the Künneth decomposition and is not a formal consequence of the definitions. It is these identities that make the evaluation map a ring homomorphism on the -algebra of formal period data (modulo bilinearity, change of variables, and Stokes). We emphasize that the surjectivity onto is tautological, since is defined in 4 as the image of this evaluation map; equivalently, (i)–(ii) are exactly what endows with its ring structure. The nontrivial statement is thus that the target is closed under multiplication with the product realized by the product period datum, not that the map hits everything.
Proof. (i) The integral is -linear in the chain and in the form by definition of integration of forms over chains; for disjoint cycles one has . This gives additivity and, applied to both slots, -bilinearity.
(ii) Write for the projections, so that is a form on . By the Fubini theorem for the product of the chains and (both of complementary dimension in their factors, so that has complementary dimension in the product), This is the analytic form of the Künneth decomposition at the level of the comparison isomorphism: the period matrix of a tensor product of motives is the Kronecker product of the period matrices, and the displayed entry is the product of the corresponding entries. This establishes the two displayed identities (i) and (ii); their ring-theoretic reading is recorded in 15. ◻
Corollary 16 (Periods respect factorized amplitudes). If a physical amplitude factorizes as with each realized by a period datum , and if the joint amplitude is realized by the product period datum , then the motivic identity realizes compatibly with the factorization. Thus the master formula of (2) is consistent with amplitude factorization at tree level and, more generally, at any kinematic factorization channel where the geometry degenerates to a product.
Proof. Immediate from 14(ii): . ◻
5 Motivic Amplitudes and Coalgebraic Anatomy
5.1 Coalgebras, comodules, and the coaction
We recall the coalgebraic vocabulary in the form we use it.
Definition 17 (Coalgebra, comodule). A coalgebra over is a -vector space with a coproduct and counit satisfying coassociativity and counitality . A right comodule over a Hopf algebra is a space with a coaction satisfying and .
Definition 18 (Hopf algebra of motivic periods). Let be a connected graded commutative Hopf algebra of motivic-Galois type: (connectedness), the grading is by weight, and the coproduct , product, unit , counit (projection onto ), and antipode make a Hopf algebra. For mixed Tate motives over a number field, is the Hopf algebra of the motivic Galois group, isomorphic as a graded vector space to the tensor coalgebra on . Write for the augmentation ideal.
Definition 19 (Motivic amplitude object). A motivic amplitude object is an element of the Hopf algebra of 18. It is equipped with the weight and depth gradings , together with the coproduct
Because is connected and graded, the reduced coproduct
Definition 20 (Cobracket). On the associated graded Lie coalgebra of indecomposables of (the “Lie coalgebra of primitives”), the cobracket is the composite
Definition 21 (Coalgebraic anatomy). Given a class of observable objects equipped with a coalgebraic operation or , the coalgebraic anatomy of is the tree (or, after antisymmetrization, the collection of Lie words) of lower-complexity pieces obtained by iterating the decomposition. The primitives are the leaves: elements with , equivalently .
5.2 The Conditional Amplitude Decomposition theorem
We now prove the central structural result: any realization channel that is monoidal transports the coalgebraic anatomy of to a genuine decomposition of the realized amplitude. This is the theorem that makes 6 of Part I concrete.
Theorem 22 (Conditional Amplitude Decomposition). Let be the connected graded Hopf algebra of motivic periods (18) with coproduct , and let be the corresponding connected graded Hopf algebra on the realization side, with coproduct . Let be the homomorphism of graded Hopf algebras induced on the motivic-Galois Hopf algebras by a Tannakian realization functor of 7; concretely is the grading-preserving -linear algebra map that intertwines the coproducts, i.e. the square
- →
- ↓
- ↓
- →
Proof. Everything takes place at the level of graded vector spaces and -linear maps; the realization functor enters only through the induced Hopf-algebra homomorphism , so there is no application of a functor to a vector-space element. Evaluate the commuting square compat on . Along the top-right path, using (5) and -bilinearity of . Along the left-bottom path, . Commutativity of compat equates the two, giving the claimed identity. Existence of as a graded Hopf-algebra homomorphism is exactly the statement that a Tannakian realization functor induces a homomorphism of the associated motivic-Galois Hopf algebras (functoriality of the Tannakian dual); the square compat is then the compatibility of that homomorphism with the coproducts, which is part of the definition of a Hopf-algebra homomorphism. Naturality in is the linearity of ; compatibility with gradings holds because is grading-preserving (weight and depth are motivic invariants transported by realization). For the period realization , the displayed identity is precisely the statement that the numerical coproduct of a period is computed by the Goncharov/Brown coaction, which is the established content of [3, 7, 10]. ◻
Remark 23 (Where the S/H/P labels bite). 22 is, as pure algebra, a formal property of Hopf-algebra homomorphisms and their compatibility with coproducts — it is always true once the hypotheses hold. Its physical content is entirely in the antecedent “”: whether a given physical amplitude actually admits a motivic lift lying in the Hopf algebra of motivic periods. For Feynman integrals evaluating to multiple zeta values or (mixed Tate) polylogarithms this is established (status ; e.g. [10]). For amplitudes requiring elliptic or higher motivic structures the comodule hypothesis fails over the Tate Hopf algebra (31); the decomposition is then only heuristic (status ) until one passes to the appropriate elliptic coalgebra. Thus 22 is a status- theorem about a status-conditional hypothesis; the conditionality is exactly the epistemic-honesty content of the S/H/P system.
5.3 Iterated coactions and the anatomy tree
Iterating the reduced coaction produces the full anatomy. Write , etc. Coassociativity ensures the iterates are well-defined up to the coalgebra axioms, and the maximal iteration terminates because weight strictly decreases in each nontrivial tensor factor.
Proposition 24 (Termination and primitivity of the anatomy). For a motivic amplitude object of weight , the iterated reduced coaction terminates after at most steps, and its terminal pieces are primitives (weight-one for the mixed Tate polylogarithmic coalgebra). Consequently the coalgebraic anatomy of is a finite tree whose leaves are primitive periods.
Proof. The reduced coaction is graded and lowers weight: if then every term of has with both factors of weight (the weight-zero part is exactly the trivial terms subtracted in ). Hence each factor has weight . By induction on weight, after at most further steps every factor reaches weight one, at which point vanishes (weight one cannot split into two positive weights summing to one). Weight-one indecomposables are the primitives; for the mixed Tate polylogarithmic coalgebra these are the logarithms (26). Finiteness of the tree follows since each node has finitely many children (the coaction sum is finite) and depth is bounded by . ◻
6 The Goncharov Lie Coalgebra
6.1 Construction
We now construct the central object of Source 2 of the knowledge base — the Goncharov Lie coalgebra of a field — and prove its generation theorem.
Definition 25 (Goncharov Lie coalgebra). Let be a field. The Goncharov Lie coalgebra is the graded Lie coalgebra of framed mixed Tate motives over , generated by the (motivic) polylogarithm classes and their multiple variants, graded by weight , and equipped with the cobracket
The cobracket is the shadow, on primitives, of Goncharov’s explicit coproduct formula for the iterated integral :
6.2 The cogeneration theorem
Theorem 26 (Weight-graded primitives cogenerate ). Let be a number field (so that the freeness properties of the motivic Galois Lie algebra of 2 hold, by Borel’s theorem and Deligne–Goncharov). Let with cobracket (8), and set (the weight- primitives). Then:
the primitives cogenerate : every element of weight is recovered from iterated cobrackets of primitives via the coradical filtration;
there is a canonical isomorphism , sending the weight-one primitive attached to to under the period map;
the weight filtration given by is exhaustive, and the associated graded is spanned by iterated cobrackets of weight-one elements.
Proof. (b) Weight-one framed mixed Tate motives over are extensions of by , classified by (the Kummer motives of 11). Weight one cannot split as with , so and ; the period of the Kummer motive attached to is , giving the stated map.
(c) The weight filtration is exhaustive by construction ( is a direct sum of its graded pieces). For the associated graded, note that is graded and, by the structure of the Goncharov coproduct (9), its leading term in the depth filtration expresses a weight- generator modulo lower depth as a sum of cobrackets of strictly lower weight. Dualizing: the graded dual of is a graded Lie algebra (the motivic Galois Lie algebra), which for a number field is a free graded Lie algebra on generators in each weight dual to the motivic -groups (freeness by Borel’s regulator theorem and Deligne–Goncharov; over an arbitrary field freeness may fail, which is why we restrict to number fields here); a free graded Lie algebra is generated by its degree-wise generators, and the cobracket structure on the dual coalgebra is therefore cogenerated by the corresponding primitives.
(a) Combine (b) and (c) with the coradical filtration of a graded connected coalgebra: for such a coalgebra the primitives cogenerate, i.e. the intersection of the kernels of all iterated reduced coactions is zero, so every element is detected by, and reconstructed from, its iterated cobrackets landing in tensor powers of primitives. By 24 the iteration terminates in weight , and by (b) the terminal primitives are . Hence the weight-graded primitives cogenerate as a graded Lie coalgebra. ◻
Remark 27 (Physical reading, status ). Under the amplitude dictionary, weight transcendentality loop/functional complexity, and the cobracket single discontinuity/cut. 26 then reads: all amplitude complexity is generated by iterated single cuts of logarithmic (weight-one) building blocks. This is a candidate rigorous counterpart of the physics folklore that multi-loop symbols are built from iterated discontinuities. We flag the physical reading as status ; the mathematical statement (a)–(c) is status .
6.3 The symbol as maximal iteration
The symbol of a weight- polylogarithm is the maximal iteration of the cobracket into weight-one pieces: . It records the leaves of the anatomy tree of 24 in order, and is computed by fully un-shuffling the coproduct (9).
Example 28 (Symbol of the dilogarithm). For the Bloch–Wigner motive of , the reduced coaction is so the symbol is . The two tensor slots are the two weight-one primitives; the cobracket is their antisymmetrization . Physically the two entries are the two branch loci and of the dilogarithm — its cut structure — exactly as the dictionary of 7 predicts.
7 Coproducts, Cuts, and Factorization
7.1 The decomposition dictionary
The coalgebraic operations acquire direct physical readings. We reproduce the relevant part of the operations dictionary with status labels.
| Coalgebraic operation | Physical interpretation | Status |
|---|---|---|
| coproduct/coaction | full anatomy: how an amplitude splits into lower pieces | |
| cobracket | antisymmetric primitive split; single cut/factorization channel | |
| Goncharov coproduct (9) | arithmetic anatomy of (multiple) polylogs | |
| symbol | maximal discontinuity data; branch-cut structure | |
| primitive element | irreducible informational contribution | |
| weight grading | transcendental complexity loop/functional depth | / |
| depth grading | iterated-integral nesting nested propagation layers | / |
| residue / pole | singular channel, threshold, on-shell condition | / |
| monodromy | analytic continuation across a cut |
7.2 Cuts and discontinuities
The physical content of the cobracket is that a single cut of an amplitude — putting an internal propagator on shell, or taking a discontinuity across a branch point — is computed by the first entry of the coaction.
Proposition 29 (Discontinuity computes the first coaction slot). Let be a motivic amplitude object with . Under the period realization, the discontinuity of across the branch point associated to a weight-one primitive (where is a kinematic invariant vanishing on the cut) is i.e. the first tensor factors paired with the second factor . Iterating recovers the full sequence of multiple discontinuities as deeper coaction slots.
Proof. The monodromy of around acts on the Betti realization by the Picard–Lefschetz transformation, whose logarithm is the weight-one nilpotent attached to . Under the motivic coaction the monodromy operator is dual to the second tensor slot: the coaction intertwines analytic continuation with the -comodule structure, so the variation picks out exactly the terms whose second factor is , with coefficient from the period of (namely ). Realizing gives the displayed formula; iterating on the first factor (itself a lower-weight amplitude object) yields the multiple discontinuities as iterated first-slots. ◻
7.3 Factorization and positive geometries
At kinematic boundaries where an amplitude factorizes, the geometry underlying the period degenerates. In the positive-geometry description [12], an amplitude is the canonical form of a positive geometry , and its boundary strata — the faces of — are themselves positive geometries whose canonical forms are the factorized (residue) amplitudes.
Example 30 (Boundary strata as coaction, status /). For the amplituhedron and associahedron picture of tree-level amplitudes, the residue of the canonical form on a codimension-one face is the product of two lower-point canonical forms. Matching this to 16: the codimension-one boundary is a product period datum , so the residue equals . The coaction’s leading term organizes the poset of boundary strata exactly as the coalgebra’s coradical filtration organizes the anatomy tree. The identification of “boundary stratum” with “coaction slot” is status where the geometry is a genuine positive geometry with known canonical form and where it is only conjectural.
7.4 The Connes–Kreimer Hopf algebra
On the renormalization side, the same coalgebraic grammar appears as the Connes–Kreimer Hopf algebra of Feynman graphs [11]: the coproduct
- →
- ↓
- ↓
- →
8 The Non-Tate Obstruction and the Cosmic Galois Group
8.1 A limitation theorem
Not every amplitude is polylogarithmic. The following theorem gives the series’ Limitation 4 (“some amplitudes require elliptic, modular, non-Tate structures”) a precise, citable form.
Theorem 31 (Non-Tate Obstruction). Let be a motivic amplitude object whose motivic lift has, as a subquotient, the motive of an elliptic curve with (or, more generally, a non-Tate simple motive). Then:
is not a comodule element over the mixed Tate Hopf algebra of iterated integrals of -forms; equivalently, its coaction does not close within the Tate coalgebra generated by ;
the weight-graded polylogarithmic anatomy of 26 is unavailable: the primitives are no longer exhausted by , and the period matrix contains the elliptic periods (and quasi-periods ) of , which are not -linear combinations of multiple zeta values;
consequently one must replace by an elliptic-polylogarithm coalgebra (iterated integrals on , or on the modular curve), and 22 holds only relative to that larger Hopf algebra.
Proof. (ii) The category of mixed Tate motives over has simple objects only the Tate twists ; their periods lie in the -algebra generated by and multiple zeta (and polylog) values. The motive is simple of rank two and not isomorphic to any (its Hodge numbers are , whereas Tate motives are of type ). Its period matrix is with (Legendre relation); by transcendence results (Chudnovsky, and the theory of periods of elliptic curves) is not a -linear combination of powers of and MZVs for without complex multiplication. Hence the elliptic periods are genuinely new.
(i) If were a comodule element over , its coaction would express all its “anatomy” using only Tate generators, forcing every subquotient of its motivic lift to be Tate — contradicting the presence of as a subquotient. Concretely, the second tensor slots of would all be logarithms/zeta values, but the elliptic period appears in with a monodromy (an modular transformation) that no Tate comodule structure can reproduce.
(iii) The correct home is the Hopf algebra of iterated integrals of modular forms / functions on (Brown, Levin–Racinet, Broedel–Duhr–Dulat–Tancredi). Over this larger algebra the comodule hypothesis of 22 is restored, and the theorem applies verbatim with replaced by . This proves (iii). ◻
Corollary 32 (Sharpness of the master formula). The master formula of (2) remains valid for elliptic amplitudes, but the coalgebraic anatomy must be taken over , not . The status of the polylogarithmic decomposition drops from to “inapplicable” precisely at the elliptic threshold; this is the honest boundary of the mixed Tate story.
8.2 The cosmic Galois group
Brown’s cosmic Galois group is a proposed pro-algebraic group acting on the ring of (motivic) periods of a QFT, compatible with the coaction [8, 9]. We record the conditional compatibility statement, carefully flagged.
Proposition 33 (Cosmic Galois compatibility, status /). Suppose a cosmic Galois group acts on the Hopf algebra of motivic periods by Hopf automorphisms, i.e. its coaction on the ring of motivic amplitude objects satisfies for . Then the action descends through the period realization to an action on the numerical amplitude ring commuting with the coaction-transported decomposition of 22. In particular the -orbit of an amplitude constant (e.g. ) is closed under the coaction, and the “small graph” / weight-drop phenomena of periods are -equivariant.
Proof (conditional/heuristic). Formally, if is a subgroup of the Tannakian automorphism group of the fibre functor acting by Hopf automorphisms, then by definition it commutes with the comodule structure , and 22 transports this equivariance through . The content is entirely in the antecedent: a precise definition of for a given QFT is open (it is conjecturally the motivic Galois group of the category of periods appearing in that theory, cut down by physical constraints such as the absence of / even weights in certain schemes). Because the antecedent is conjectural, the whole statement carries status /, matching the knowledge base’s labeling of the cosmic Galois entry. The verified instances — the coaction on periods up to high loop order [10] — provide status- evidence for special cases but do not establish the general . ◻
Remark 34. 33 is deliberately the most speculative result in the paper. It is included because the cosmic Galois group is the natural “symmetry organizing amplitudes” that the whole coalgebraic picture points toward; but per the discipline of 1.3 we do not present it as established. It sets up Part III (variation of the geometry over moduli, where -equivariance becomes a constraint on Gauss–Manin connections) and Part VI (Hopf-algebraic renormalization, where acts on the Connes–Kreimer side).
9 Results
We collect the principal results of the paper and their epistemic status.
14 (Multiplicativity, status ). The period pairing is additive and multiplicative: , so the period map is a ring homomorphism onto the Kontsevich–Zagier period ring. This licenses to respect factorization (16).
22 (Conditional Amplitude Decomposition, status on a conditional hypothesis). Any realization channel, as a graded Hopf-algebra homomorphism compatible with the coproduct, transports to a decomposition of the realized amplitude. This is the concrete form of Part I’s Decomposition Axiom (6). The physical content is in whether a given amplitude admits a motivic lift in (23).
24 (Termination, status ). The iterated coaction of a weight- amplitude object terminates in at most steps at primitive leaves.
26 (Cogeneration, status ). The weight-graded primitives cogenerate the Goncharov Lie coalgebra , and weight-one primitives are (logarithms). Physically (status ): all amplitude complexity is built from iterated cuts of logarithms.
29 (Discontinuity, status ). The discontinuity of an amplitude across a branch point is the first coaction slot paired with the corresponding weight-one primitive.
31 (Non-Tate Obstruction, status ). Elliptic (non-Tate) motivic lifts escape the mixed Tate coalgebra; the polylogarithmic anatomy is unavailable and one must pass to an elliptic-polylogarithm Hopf algebra. This is the rigorous form of the series’ Limitation 4.
33 (Cosmic Galois, status /). Conditional on a precise definition of the cosmic Galois group, its action commutes with the coaction and descends to the numerical amplitude ring. Presented as conjectural.
The logical dependency of the results is displayed below.
- ↓
- ↘
- ↓
- ↙
- ↓
- ↓
- ↓
- ↙
10 Formalization: Haskell and Lean
To honor the Curry–Howard–Lambek bridge of the series (propositions as types, proofs as programs, realization as compilation), we accompany the mathematics with a compiling Haskell formalization of the coalgebra/coproduct structure and an idiomatic Lean type sketch. The full sources live in src/motives-periods-amplitudes/ and lean/motives-periods-amplitudes/; we display the load-bearing fragments.
10.1 Haskell: the coalgebra and coproduct
The Haskell package represents a graded coalgebra element as a formal -linear combination of words in generators, with the Goncharov-style deconcatenation coproduct. Weight is the word length (for the shuffle/deconcatenation model) and the reduced coproduct strips the two trivial terms; primitivity is decidable.
-- Word in generators; weight = length; deconcatenation coproduct
newtype Word' g = Word' [g] deriving (Eq, Ord, Show)
-- Formal Q-linear combination (rationals as coefficients)
newtype LinComb g = LinComb (Map (Word' g) Rational)
-- Deconcatenation coproduct: Delta(w) = sum over splittings u | v
coproduct :: Ord g => Word' g -> [(Word' g, Word' g)]
coproduct (Word' xs) =
[ (Word' (take k xs), Word' (drop k xs)) | k <- [0 .. length xs] ]
-- Reduced coproduct: drop the two trivial terms (empty | w) and (w | empty)
reducedCoproduct :: Ord g => Word' g -> [(Word' g, Word' g)]
reducedCoproduct w@(Word' xs) =
[ p | p@(Word' a, Word' b) <- coproduct w, not (null a), not (null b) ]
-- Cobracket delta = antisymmetrization of the reduced coproduct.
-- NOTE: this returns a formal signed sum WITHOUT combining like terms;
-- algebraic simplification in the exterior algebra Lambda^2 L is omitted
-- (adequate for the demonstration, not a normal form).
cobracket :: Ord g => Word' g -> [(Word' g, Word' g, Rational)]
cobracket w =
[ (a, b, 1) | (a, b) <- reducedCoproduct w ] ++
[ (b, a, -1) | (a, b) <- reducedCoproduct w ]
-- Primitivity: delta(x) = 0 <=> reduced coproduct empty <=> weight <= 1
isPrimitive :: Ord g => Word' g -> Bool
isPrimitive w = null (reducedCoproduct w)
The demonstration program verifies, on the dilogarithm symbol of 28, that (a) coassociativity holds numerically for the deconcatenation coproduct, (b) weight-one words are exactly the primitives (26(b)), and (c) the iterated reduced coproduct of a weight- word terminates after steps at weight-one leaves (24). It compiles with ghc and runs to print the anatomy tree of a sample amplitude word.
10.2 Lean: period, amplitude, and coalgebra types (best-effort sketch)
The Lean file records the type signatures of the objects and the statement of 22 as a definition-level goal (with sorry placeholders, as a non-build-gated sketch).
structure PeriodDatum (X Form Cycle : Type) where
space : X
divisor : X
form : Form
cycle : Cycle
-- per : PeriodDatum -> C (integration pairing, abstracted)
def per {X F C : Type} (integrate : C -> F -> Complex)
(Pi : PeriodDatum X F C) : Complex := integrate Pi.cycle Pi.form
-- A right H-comodule of motivic amplitude objects
structure MotivicAmplitude (H A : Type) where
coaction : A -> List (A x H) -- Delta(a) = sum a_{i,1} (x) a_{i,2}
weight : A -> Nat
depth : A -> Nat
-- Cobracket as antisymmetrized reduced coaction on primitives L
def cobracket {L : Type} (coact : L -> List (L x L)) (x : L) :
List (L x L) := coact x -- (antisymmetrization elided in sketch)
-- T1 (Conditional Amplitude Decomposition), signature-level.
-- The realization is a MAP between carrier types (a Hopf-algebra
-- homomorphism), NOT a functor applied to elements: this is exactly the
-- corrected formulation of Theorem 5.3.
theorem coaction_transports
{H A R : Type} (ma : MotivicAmplitude H A)
(Real : A -> R) (RealH : H -> R) (target : R -> List (R x R)) (a : A) :
target (Real a)
= (ma.coaction a).map (fun p => (Real p.1, RealH p.2)) := by
sorry
The Lean sketch is explicitly best-effort and not build-gated; it fixes the type-level anatomy (period data, comodule coactions, cobrackets) so that a future Mathlib-backed development can discharge the sorrys. Note that modelling the realization as an honest map Real : A -> R between carrier types (rather than a functor applied to elements) is precisely the type-level reflection of the Hopf-algebra-homomorphism formulation of 22; the period_multiplicative statement in the same file is proved outright (no sorry) from the Fubini/Künneth hypothesis.
11 Discussion
11.1 Limitations (reproduced and addressed)
Per the series discipline, we restate the global limitations and indicate how Part II addresses each.
Not every mathematical analogy is a physical theory. The S/H/P labels are part of the formalism (1.3); the strongest results here (14, 22, 26) are status as mathematics, while their physical readings are labelled .
Motives are not literally “the functions of” an object. Addressed head-on in the boxed 10; the functional-essence reading is status throughout.
No single universal functor . Part II defines only the motivic channel , , over the domain of motivic period theory — a local, domain-indexed construction, not a global functor.
Not every observable is a period. Made precise as the Non-Tate Obstruction theorem (31): elliptic and modular amplitudes require larger coalgebras.
Gauge redundancy physical symmetry. Deferred to Part VI, but foreshadowed in 1: the dashed equivalence is a motivic (gauge/duality) equivalence realizing the same amplitude.
11.2 Bridges to Parts I and III
Part II builds on Part I by instantiating its abstract pipeline: , the cohomological realizations, , and 6 the Goncharov coproduct. It sets up Part III by handing off the motivic amplitude objects and their families: as kinematic parameters vary, the variety in a period datum moves in a moduli space, the period becomes a multivalued function on that base, and the coaction becomes a statement about the Gauss–Manin/Picard–Fuchs connection governing the variation of Hodge structure. In the ladder’s language, Part III lets Part II’s motives “live over and vary across concrete geometric moduli.”
- →
- →
11.3 The four slogans
The series’ compressed theses read, in the light of Part II: mathematics is the syntax of physical representation; motives are functional semantics (9, status ); periods are numerical measurements (13, 14); coalgebras are decomposition laws (22, 26). Part II is precisely the module that turns the last two slogans into theorems.
12 Conclusion
We have given the motivic channel of the MathPhysics Representation Library its first fully concrete instantiation. Motives supply the pre-numerical functional essence (status , carefully distinguished from the mathematical definition); the four cohomological realizations supply Part I’s abstract channels ; and the period map supplies the observable. On this foundation the master formula with its coaction and cobracket becomes a working piece of mathematics: the period pairing is multiplicative (14), the coaction transports through realizations (22), the Goncharov Lie coalgebra is cogenerated by its weight-one logarithmic primitives (26), amplitude discontinuities are read off the first coaction slot (29), and the whole polylogarithmic anatomy has a sharp boundary at the elliptic threshold (31). The most speculative extension — the cosmic Galois symmetry (33) — is flagged as such and handed forward.
The emergent property Part II adds to the ladder is the internal coalgebraic anatomy of observables: the discovery that a physical amplitude is not an atom but carries a canonical decomposition into primitive informational pieces, governed by a Lie coalgebra. In the next module this anatomy is set in motion over the moduli spaces of algebraic geometry; in the capstone module the same Hopf-coalgebraic grammar returns as the bookkeeping of gauge redundancy and renormalization. Coalgebras, in short, are the grammar by which physics takes its observables apart.
[1] M. Kontsevich and D. Zagier, Periods, in “Mathematics Unlimited — 2001 and Beyond,” Springer, 2001, pp. 771–808.
[2] A. B. Goncharov, Multiple polylogarithms and mixed Tate motives, arXiv:math/0103059 (2001).
[3] A. B. Goncharov, Galois symmetries of fundamental groupoids and noncommutative geometry, Duke Math. J. 128 (2005), no. 2, 209–284; arXiv:math/0208144.
[4] A. B. Goncharov, M. Spradlin, C. Vergu, and A. Volovich, Classical polylogarithms for amplitudes and Wilson loops, Phys. Rev. Lett. 105 (2010), 151605; arXiv:1006.5703.
[5] S. Bloch, H. Esnault, and D. Kreimer, On motives associated to graph polynomials, Comm. Math. Phys. 267 (2006), 181–225; arXiv:math/0510011.
[6] F. Brown, Mixed Tate motives over , Annals of Mathematics 175 (2012), no. 2, 949–976; arXiv:1102.1312.
[7] F. Brown, On the periods of some Feynman integrals, arXiv:0910.0114 (2009).
[8] F. Brown, Feynman amplitudes, coaction principle, and cosmic Galois group, Commun. Number Theory Phys. 11 (2017), 453–556; arXiv:1512.06409.
[9] F. Brown, Periods and Feynman amplitudes, arXiv:1512.09265 (2015).
[10] E. Panzer and O. Schnetz, The Galois coaction on periods, Commun. Number Theory Phys. 11 (2017), 657–705; arXiv:1603.04289.
[11] A. Connes and D. Kreimer, Renormalization in quantum field theory and the Riemann–Hilbert problem I: the Hopf algebra structure of graphs and the main theorem, Comm. Math. Phys. 210 (2000), 249–273; arXiv:hep-th/9912092.
[12] N. Arkani-Hamed, Y. Bai, and T. Lam, Positive geometries and canonical forms, JHEP 11 (2017) 039; arXiv:1703.04541.
[13] P. Deligne and A. B. Goncharov, Groupes fondamentaux motiviques de Tate mixte, Ann. Sci. École Norm. Sup. 38 (2005), 1–56.
[14] F. Brown, On the decomposition of motivic multiple zeta values, in “Galois–Teichmüller theory and arithmetic geometry,” Adv. Stud. Pure Math. 63 (2012), 31–58; arXiv:1102.1310.
[15] J. Broedel, C. Duhr, F. Dulat, and L. Tancredi, Elliptic polylogarithms and iterated integrals on elliptic curves, JHEP 05 (2018) 093; arXiv:1712.07089.
[16] C. Duhr, Mathematical aspects of scattering amplitudes, in “Journeys Through the Precision Frontier: Amplitudes for Colliders” (TASI 2014), World Scientific, 2016, pp. 419–476; arXiv:1411.7538.
13 The Motivic and Amplitude Library
For reference we reproduce the motivic/amplitude dictionary of the series, with epistemic status labels retained verbatim. This table is the Appendix B of the source library, restricted to Part II’s domain.
| Mathematics | Physical representation | Decomposition meaning | Status |
|---|---|---|---|
| Field | Kinematic/coordinate domain | Allowed parameter values | / |
| Nonzero scales or ratios | Energies, masses, cross-ratios | / | |
| Cross-ratio | Conformal invariant | Dimensionless observable | |
| Algebraic variety | Constraint solution space | Allowed configurations | |
| Graph hypersurface | Geometry of a Feynman graph | Container of graph integral data | |
| Feynman integral | Period integral | Observable from geometry domain | |
| Period | Measurable number from geometry | Numerical shadow of structure | |
| Period pairing | Amplitude contribution | Pairing of form and cycle | |
| Differential form | Integrand / observable density | What is accumulated | |
| Cycle | Integration domain / history class | Where accumulation occurs | |
| Relative cohomology | Boundary-sensitive observable | Cuts, thresholds, boundaries | |
| Motive | Pre-numerical functional essence | Hidden source of periods | / |
| Mixed motive | Layered informational object | Multiple weights/depths | / |
| Mixed Tate motive | Polylogarithmic period structure | Tame amplitude period class | |
| Non-Tate motive | More complex amplitude geometry | Elliptic, Calabi–Yau, modular | |
| Multiple zeta value | Special period | Constant term in loop expansions | |
| Polylogarithm | Layered propagation function | Iterated contribution | |
| Multiple polylogarithm | Multi-scale iterated propagation | Nested channel dependence | |
| Symbol of polylogarithm | First-order function decomposition | Branch-cut/discontinuity data | |
| Coproduct / coaction | Decomposition of a period | How amplitude splits | |
| Cobracket | Antisymmetric primitive split | Factorization/information split | |
| Goncharov coproduct | Motivic decomposition operation | Arithmetic anatomy of polylogs | |
| Goncharov Lie coalgebra | Coalgebra of polylog data | Grammar of primitive amplitude data | / |
| Weight grading | Transcendental complexity | Loop depth/functional complexity | / |
| Depth grading | Iterated-integral nesting | Nested propagation layers | / |
| Primitive element | Indecomposable motivic component | Irreducible contribution | |
| Regulator map | Motive to analytic number | Measurement/numerical extraction | / |
| Hodge realization | Analytic-geometric reading | Differential-form representation | |
| de Rham realization | Form/integrand side | Calculational representation | |
| Betti realization | Cycle/topological side | Path/integration-domain rep. | |
| Étale realization | Arithmetic/discrete realization | Number-theoretic shadow | / |
| Cosmic Galois group | Symmetry acting on periods | Hidden symmetry of amplitudes | / |