Elliptic Double Shuffle Equations

• Double Shuffle Equations are algebraic relations that characterize elliptic multizeta values by combining shuffle and stuffle products with arithmetic corrections modulo 2πi.
• They utilize iterated Eisenstein integrals and mould theory to translate geometric and arithmetic features into a unified algebraic framework.
• This framework establishes a complete system of generators and relations for elliptic multizetas, paralleling classical multiple zeta value theory.

Elliptic double shuffle equations refer to a family of algebraic relations governing the structure of elliptic multizeta values (EMZs)—functions on the upper-half plane $\mathfrak{H}$ generated by coefficients in the elliptic associator, itself arising as Enriquez's genus-one analogue of the Drinfel'd associator. These equations extend the double shuffle framework familiar from the classical theory of multiple zeta values (MZVs) by incorporating geometric and arithmetic features specific to the elliptic setting, notably via iterated Eisenstein integrals and arithmetic corrections modulo $2\pi i$. The central result is that, under suitable conditions, the elliptic double shuffle relations characterize all algebraic dependencies among EMZs, paralleling the role of double shuffle relations for classical multiple zeta values (Lochak et al., 2017).

1. Formalism of Elliptic Multizeta Generating Series

Let $E(\tau)$ denote the Lie-like elliptic generating series,

$$E(\tau) = \log C(\tau) \in ({}^{\text{geom}} \otimes \mathbb{Z}[2\pi i])\langle\!\langle a, b \rangle\!\rangle,$$

where $C(\tau) = g(\tau) \cdot C$, with $g(\tau) \in \exp(\text{Der}_0(\,\_2))$ constructed from iterated Eisenstein integrals ("geometric part") and $C = \exp E$, $E$ derived from genus 0 associator data. The expansion in noncommuting monomials

$$E(\tau) = \sum_{r \geq 1} \sum_{k_1,\ldots,k_r \geq 1} E(k_1,\ldots,k_r;\tau)\, c_{k_1}\cdots c_{k_r},$$

with $c_k = \text{ad}(a)^{k-1}(b)$, defines the coefficients $E(k_1,\ldots,k_r;\tau)$ as elliptic multizetas (EMZs).

To obtain a commutative presentation, one passes through a "mould-isomorphism": $$ma: c_{k_1}\cdots c_{k_r} \mapsto (-1)^{\sum k_i - r} u_1^{k_1-1}\cdots u_r^{k_r-1},$$ and then sets $u_i = X$, introducing the tracking variable $Y^r$, yielding the series

$$E(\tau; X, Y) = \sum_{r \geq 1} \sum_{n_1,\ldots,n_r \geq 1} \zeta^{\mathrm{ell}}(n_1,\ldots,n_r;\tau) X^{n_1-1}\cdots X^{n_r-1} Y^r.$$

This construction links geometric objects (iterated Eisenstein integrals), arithmetic corrections, and their algebraic representations in both noncommutative and mould-theoretic frameworks.

2. Double Shuffle Structure: Shuffle and Stuffle Products

The elliptic double shuffle system comprises two families of relations, each corresponding to different product structures on EMZs:

• Shuffle relations: Stemming from the group-likeness of $C(\tau) = \exp(E(\tau))$, these reflect the behavior of the shuffle product under concatenation of iterated integrals. In the commutative variable formalism,

$E(\tau; X_1, Y) \, \shuffle \, E(\tau; X_2, Y) = E(\tau; X_1 + X_2, Y).$

This yields specific linear dependencies among EMZs indexed by shuffle permutations.

• Stuffle ("quasi-stuffle") relations: These arise from the exponential behavior of the depth-tracking variable $Y$, incorporating corrections involving powers of $2\pi i$,

$$E(\tau; X, Y_1) * E(\tau; X, Y_2) = E(\tau; X, Y_1 + Y_2) + \text{corrections}.$$

Modulo $2\pi i$, these corrections vanish, and the relations simplify to

$$\overline{E}(\tau; X, Y_1) * \overline{E}(\tau; X, Y_2) = \overline{E}(\tau; X, Y_1 + Y_2) \pmod{2\pi i}.$$

In mould-theoretic language, the system requires that the depth-$r$ component of the mould $\Delta^{-1}(\mathfrak e_m(\tau))$ be both alternal and have swap-alternality ("bialternality").

3. Algebraic Structure: Decomposition and Lie Theory

Working modulo $2\pi i$, the space generated by EMZs decomposes into geometric and arithmetic parts. Specifically, the algebra generated by mod-$2\pi i$ EMZs and $2\pi i\tau$ is isomorphic to

$${}^{\text{geom}} \otimes_{\mathbb{Q}} \overline{\mathbb{Q}} \cong U(\text{Der}_0(_2))^\vee,$$

with ${}^{\text{geom}}$ the Hopf algebra of iterated Eisenstein integrals. The arithmetic component, built from the (mod $2\pi i$) EMZs associated to the "arithmetic" subseries, also embeds into this tensor product after adjoining $2\pi i\tau$.

On the Lie algebra level, the elliptic double shuffle Lie algebra $\mathfrak{ds}_{\rm ell}$ is characterized by

$$\mathfrak{ds}_{\rm ell} = \{\, f \in \text{Lie}\langle\!\langle a, b \rangle\!\rangle : ma(f) \in \text{ARI}^{\Delta\text{-}al*al} \,\},$$

with

$$0 \longrightarrow \mathfrak{r}_{\rm ell} \longrightarrow \mathfrak{ds}_{\rm ell} \longrightarrow \mathfrak{ds} \longrightarrow 0$$

a short exact sequence of Lie algebras, and with $$\gamma\colon \mathfrak{ds} \to \mathfrak{ds}_{\rm ell}$$ a splitting. $\mathfrak{ds}$ is the classical double shuffle Lie algebra, while $\mathfrak{r}_{\rm ell}$ captures the genus-one "new" kernel.

4. Completeness of the Elliptic Double Shuffle Relations

The main conceptual result is conditional completeness: if

• (a) classical double shuffle relations generate all algebraic relations among classical MZVs modulo $2\pi i$, and
• (b) the semidirect splitting $$\mathfrak{ds}_{\rm ell} \cong \mathfrak{r}_{\rm ell} \rtimes \gamma(\mathfrak{ds})$$ holds (as established for the elliptic Grothendieck–Teichmüller Lie algebra by Enriquez),

then

$$U(\mathfrak{ds}_{\rm ell})^{\vee} \cong {}^{\text{geom}} \otimes \overline{\mathbb{Q}} \cong \mathbb{Q}\langle\!\langle \text{EMZs} \rangle\!\rangle,$$

so the elliptic double shuffle relations suffice to generate all algebraic relations among EMZs modulo $2\pi i$ (Lochak et al., 2017).

5. Context, Connections, and Significance

Elliptic double shuffle theory generalizes the motivic and combinatorial structures of classical MZVs to genus one. Major components include Enriquez's elliptic associator, the role of iterated Eisenstein integrals, and the identification of alternality and swap-alternality conditions through mould theory. By replicating key features of the genus-zero motivic Galois group and its Lie algebra in the elliptic setting, this theory gives a uniform algebraic and geometric context for the study of elliptic periods.

If the conjectural completeness and semidirect decomposition are validated, the framework would offer a full set of algebraic generators and relations for $\mathbb{Q}$-algebras of EMZs and their reductions, mirroring known results for MZVs and giving new geometric and arithmetic invariants in modular and Galois-theoretic settings.

6. Further Directions and Conditionalities

Key open questions include confirming that classical double shuffle relations are exhaustive (for MZVs), establishing categorical and motivic analogues for the elliptic setting, and rigorously proving the required semidirect decomposition for $\mathfrak{ds}_{\rm ell}$. The precise arithmetic structure and the interplay with period conjectures (as in genus-zero) are active areas, with Enriquez's results on associators and the motivic Galois group providing models for future progress.

The modular features and new "elliptic kernels" suggest extensions to $p$-adic, quantum, and categorical frameworks, and potential applications in arithmetic geometry, deformation theory, and mathematical physics.