The Schwinger–Dyson Equations and the Ward–Takahashi Identities in Biquaternionic Form

Introduction

The classical equations of motion are the statement that the action is stationary. Two exact relations of the quantum theory descend from that statement by one and the same device — the integration of a total derivative in field space — and they are the subject of this article.

The Schwinger–Dyson equations are the quantum transcription of the Euler–Lagrange equations: the variation of the action vanishes inside correlation functions, up to contact terms generated by the variation of the operators inserted at coincident points. The Ward–Takahashi identities are the quantum transcription of Noether's theorem: if the functional measure is invariant under a one-parameter symmetry, then the divergence of the associated current vanishes inside correlation functions, again up to contact terms. The two are the same argument applied to two different functionals — the generic functional in the first case, the symmetry generator in the second — and both rest on a single premise, the translation invariance of the functional measure, which does not hold in general and whose failure is the anomaly.

The framework's corpus already contains both objects in fragments. The Yang–Mills Path Integral and the Faddeev–Popov Procedure in Biquaternionic Form derives the Schwinger–Dyson statement of the quantum Yang–Mills equation, $\langle(D_\mu F^{\mu\nu})^a\rangle=\langle J^{\nu a}_{\mathrm{gf}}\rangle+\langle J^{\nu a}_{\mathrm{gh}}\rangle$, and Noether's Theorem in Biquaternionic Form derives the classical current of the central phase. What is missing is the general statement: the master Schwinger–Dyson equation from which every such expectation-value equation follows, and the systematic Ward–Takahashi derivation of the current identity as the quantum Noether theorem. This article supplies that general statement and records what the algebra contributes to it.

The contribution is a reading, and it is small.

Established (algebra). The functional derivative $\delta/\delta\tilde\Phi$ is defined through the real bilinear form $\langle\tilde Q,\tilde Y\rangle=\mathrm{Re}\,\mathrm{Tr}(\tilde Q^{*}\tilde Y)$, so it is the module's own gradient; the derivation of the master equation requires no matrix basis and no coordinate chart. The variation of the quadratic action is the operator $\tilde K\tilde\Phi$ with $\tilde K=\Box-m^2$, so the Schwinger–Dyson hierarchy is generated by the central d'Alembertian and the mass term, and for the Dirac field the mass term is the right multiplication $\tilde\Psi_R\mapsto m\tilde\Psi_L$, which makes the fermionic hierarchy chirality-off-diagonal.
Established (algebra). The one-parameter symmetry whose Ward–Takahashi identity the algebra canonically carries is the central phase $\tilde\Phi\mapsto e^{i\alpha}\tilde\Phi$. Its generator is the central imaginary $i$, its Noether current lies in the material sector $\mathbb{M}_-$ (as Noether's Theorem in Biquaternionic Form establishes), and the quantum identity is the statement that this current is conserved inside correlators. The algebra supplies the symmetry and the home of its current; it does not supply the invariance of the measure.
Standard, and transcribed. The total-derivative argument itself, the assumption of translation invariance of the measure, the contact terms, the master equation, the local Ward–Takahashi derivation, the axial anomaly, and the Slavnov–Taylor identities of the non-abelian theory. The algebra supplies none of these.
Not supplied. The functional measure and its invariance. Every identity in this article is conditional on a premise the algebra cannot formulate, and the condition is stated rather than closed. This is the same gap as the measure gap of The Functional Integral in Biquaternionic Form, seen where it has its sharpest consequence.

The article proceeds as follows. A section fixes the total-derivative argument on which both relations rest. Two sections derive the Schwinger–Dyson equations — the quantum equation of motion and the master equation — and specialise them to the scalar field, the gauge field, and the fermion. A section derives the Ward–Takahashi identities in general form, and two sections treat the central phase, which is the framework's canonical case, and the axial identity with its anomaly. A section separates what the algebra supplies from what it imports, and the article closes with open questions.

Conventions. We use those of the companion articles, in particular The Functional Integral in Biquaternionic Form and The Generating Functional and the Effective Action in Biquaternionic Form. The algebra is $\mathbb{B}=\mathbb{C}\otimes_\mathbb{R}\mathbb{H}$ with basis $e_0=1,e_1,e_2,e_3$, $e_k^2=-e_0$, and central imaginary $i$. The sectors are $\mathbb{M}_-$ (anti-Hermitian, material) and $\mathbb{M}_+$ (Hermitian, informational), with $\mathbb{B}=\mathbb{M}_-\oplus\mathbb{M}_+$ and $\flat=-{}^{*}$ the anti-Hermitian conjugation. The real bilinear form is $$ \langle \tilde Q,\tilde Y\rangle = \mathrm{Re}\,\mathrm{Tr}\big(\tilde Q^{*}\tilde Y\big), $$ positive definite on $\mathbb{M}_+$ and negative definite on $\mathbb{M}_-$; the trace is $\mathrm{Tr}(\tilde P\tilde H)=2\,\mathrm{Sc}(\tilde P\tilde H)$ with $\mathrm{Tr}(e_0)=2$. The material coordinate is $\tilde Q=ict\,e_0+\mathbf{x}$, the $ict$ metric is $\eta=\mathrm{diag}(-1,+1,+1,+1)$, the gradient is $\tilde\nabla=e_0\partial_{ict}+e_1\partial_x+e_2\partial_y+e_3\partial_z$, and the d'Alembertian is $\Box=\tilde\nabla\tilde\nabla^{\natural}=\partial_{ict}^2+\Delta$; the quadratic operator is $\tilde K=\Box-m^2$. The functional integral with source is $$ Z[\tilde J]=\int\mathcal{D}\tilde\Phi\;e^{\,iS[\tilde\Phi]/\hbar+\langle\tilde J,\tilde\Phi\rangle}, $$ with $\tilde J$ valued in the module; the connectivity of $\tilde J$ with The Generating Functional and the Effective Action in Biquaternionic Form is recorded there, and the source convention $+\langle\tilde J,\tilde\Phi\rangle$ — a real pairing and no explicit $i$ — is the one used throughout the series.

The Total-Derivative Argument

Both relations come from one identity, and it is worth having it before either is named.

Let $\tilde\Phi$ be a biquaternion field, $\mathcal{D}\tilde\Phi$ the functional measure, and $F[\tilde\Phi]$ any functional of the field for which the objects below exist. Consider the integral of a total derivative in field space, $$ \int\mathcal{D}\tilde\Phi\;\frac{\delta}{\delta\tilde\Phi(x)}\Big(F[\tilde\Phi]\;e^{\,iS[\tilde\Phi]/\hbar}\Big). $$ If the measure is translation invariant — if $\mathcal{D}(\tilde\Phi+\tilde\epsilon)=\mathcal{D}\tilde\Phi$ for every shift $\tilde\epsilon$ vanishing at the boundary — then this integral is a shift of integration variable and vanishes: $$ 0=\int\mathcal{D}\tilde\Phi\;\frac{\delta}{\delta\tilde\Phi(x)}\Big(F[\tilde\Phi]\;e^{\,iS[\tilde\Phi]/\hbar}\Big). $$

Performing the derivative and taking the expectation value in the presence of a module-valued source $\tilde J$ gives the working identity $$ \Big\langle\frac{\delta F}{\delta\tilde\Phi(x)}\Big\rangle_{\!J} +\frac{i}{\hbar}\Big\langle F\,\frac{\delta S}{\delta\tilde\Phi(x)}\Big\rangle_{\!J} +\Big\langle F\,\tilde J(x)\Big\rangle_{\!J}=0, $$ where the three terms are the variation of the inserted functional, the variation of the action, and the source. The entire content of the Schwinger–Dyson and Ward–Takahashi relations is the specialisation of this identity.

The two specialisations. The first is at generic $F$: it gives the equations of motion inside correlators, and it is the Schwinger–Dyson family. The second is at $F$ built from the symmetry generator: it gives the current identity, and it is the Ward–Takahashi family. The distinction is which variation is nonzero.

The measure premise is where the algebra is silent. Translation invariance of $\mathcal{D}\tilde\Phi$ is an assumption about an infinite-dimensional measure that the finite-dimensional algebra does not carry. The algebra's own object is the field value space $\mathbb{B}$, and the measure is a product of Lebesgue measures on the coefficients at each point (as The Functional Integral in Biquaternionic Form records); translation invariance of that product is a property of the analytic regularisation, not of $\mathbb{B}$. Where it holds, the identities below hold; where it fails, the failure is carried by the anomaly, which is the subject of Anomalies and Anomaly Cancellation in Biquaternionic Form and The Fujikawa Method and the Path-Integral Anomaly in Biquaternionic Form. This article records the conditional statement and uses it.

The functional derivative is the module gradient. In the framework the derivative $\delta/\delta\tilde\Phi(x)$ is defined through the pairing $\langle\cdot,\cdot\rangle$, exactly as the source derivative of the generating functional is: it takes a functional to the module-valued function that reproduces the first variation, $$ \delta F=\Big\langle\frac{\delta F}{\delta\tilde\Phi},\delta\tilde\Phi\Big\rangle . $$ No component index and no privileged basis enter, and the variation of the action is computed in the algebra's own operations. This is the one place in the argument where the algebra is used rather than transcribed.

The Schwinger–Dyson Equations

The Quantum Equation of Motion

Set $F=1$ in the working identity. The derivative term drops and $$ \Big\langle\frac{\delta S}{\delta\tilde\Phi(x)}\Big\rangle_{\!J}=i\hbar\,\tilde J(x), $$ which at $\tilde J=0$ is the quantum equation of motion, $$ \Big\langle\frac{\delta S}{\delta\tilde\Phi(x)}\Big\rangle=0 . $$ The classical equation is the statement that the integrand of the functional integral is stationary; the quantum statement is that the expectation value of the variation vanishes. The difference is not a correction but a change of object: the classical equation is an equation between fields, the quantum one an equation between expectation values, and it holds inside the integral, where the field is integrated over.

Contact terms. Set $F=\tilde\Phi(y_1)\cdots\tilde\Phi(y_n)$ in the working identity. The first term is the variation of the product, $$ \frac{\delta F}{\delta\tilde\Phi(x)}=\sum_{k=1}^{n}\tilde\Phi(y_1)\cdots\widehat{\tilde\Phi(y_k)}\cdots\tilde\Phi(y_n)\;\delta(y_k-x), $$ a sum of products with one field removed and a delta at the point of variation. The working identity becomes $$ \Big\langle\frac{\delta S}{\delta\tilde\Phi(x)}\,\tilde\Phi(y_1)\cdots\tilde\Phi(y_n)\Big\rangle_{\!J} =i\hbar\Big(\tilde J(x)\big\langle\tilde\Phi(y_1)\cdots\tilde\Phi(y_n)\big\rangle_{\!J}-\sum_{k}\delta(y_k-x)\big\langle\cdots\widehat{\tilde\Phi(y_k)}\cdots\big\rangle_{\!J}\Big), $$ which is the hierarchy: the divergence of the current-like object $\delta S/\delta\tilde\Phi$ is a source term plus contact terms, and the contact terms are what distinguish the quantum equation from the naive insertion of the classical equation. At coincident points the equation is distributional, and the delta terms are the whole of the difference. This is general and standard; it is not special to the gauge field, and the same structure appears for the scalar and the fermion below.

The Master Equation and the Hierarchy

The expectation-value form is the physical one, but there is a compact generating form that produces the whole hierarchy at once. Because inserting $\tilde\Phi(x)$ under the functional integral is the same as applying $\delta/\delta\tilde J(x)$ to $Z[\tilde J]$, $$ \delta/\delta\tilde J(x)\;Z[\tilde J]=\int\mathcal{D}\tilde\Phi\;\tilde\Phi(x)\,e^{\,iS/\hbar+\langle\tilde J,\tilde\Phi\rangle}, $$ the working identity with generic $F$ is equivalent to the operator equation $$ \boxed{\;\Big(\frac{\delta S}{\delta\tilde\Phi}\Big)\Big[\frac{\delta}{\delta\tilde J}\Big]\,Z[\tilde J]=i\hbar\,\tilde J(x)\,Z[\tilde J]\;} $$ the master Schwinger–Dyson equation. Setting $F=1$ recovers the quantum equation of motion; expanding both sides in powers of $\tilde J$ and using $Z=\exp(W/i\hbar)$ generates the whole hierarchy of the connected correlation functions. The bracket in the box means that the classical variation $\delta S/\delta\tilde\Phi$ is evaluated with the field replaced by the operator $\delta/\delta\tilde J$.

Verification (free Gaussian). For the Euclidean Gaussian $Z_0[\tilde J]=\int\mathcal{D}\tilde\Phi\,e^{-\frac12\langle\tilde\Phi,\tilde K\tilde\Phi\rangle+\langle\tilde J,\tilde\Phi\rangle}$ one has $\delta S/\delta\tilde\Phi=\tilde K\tilde\Phi$, so the master equation reads $\tilde K\,\delta Z_0/\delta\tilde J=i\hbar\tilde J Z_0$, equivalently $\tilde K\langle\tilde\Phi\rangle_{\!J}=i\hbar\tilde J$. The finite-dimensional instance was checked with the Hermitian positive-definite $$ K=\begin{pmatrix}2&0.3&0\\0.3&1.4&0.2\\0&0.2&1.1\end{pmatrix}, \qquad J=(0.5,-0.4,0.9), $$ for which $K^{-1}J=(0.322647,-0.484316,0.906239)$ and $K\,(K^{-1}J)=J$ to machine precision; the two-point contact term $\tilde K\langle\tilde\Phi\tilde\Phi^{\mathsf T}\rangle_{\!J}=\tilde I+\tilde J\langle\tilde\Phi\rangle^{\mathsf T}_{\!J}$ was checked on the same matrix and closed with the same accuracy. The sign recorded in the box is the one that makes the $J$-expansion reproduce the free propagator $\tilde K^{-1}$ and not its negative.

The Equations in the Framework's Theories

The Scalar Field

For the central-scalar action $$ S[\tilde\Phi]=\int d^4x\;\Big\langle\tilde\Phi,\tilde K\tilde\Phi\Big\rangle+S_{\mathrm{int}}[\tilde\Phi], \qquad \tilde K=\Box-m^2 , $$ the variation is $$ \frac{\delta S}{\delta\tilde\Phi}=2\tilde K\tilde\Phi+\frac{\delta S_{\mathrm{int}}}{\delta\tilde\Phi}, $$ and the quantum equation of motion is $$ 2\big\langle\tilde K\tilde\Phi\big\rangle+\Big\langle\frac{\delta S_{\mathrm{int}}}{\delta\tilde\Phi}\Big\rangle=0, $$ equivalently, since $\tilde K$ is central and the expectation is linear, $$ \big(\Box-m^2\big)\,\big\langle\tilde\Phi\big\rangle=-\tfrac12\Big\langle\frac{\delta S_{\mathrm{int}}}{\delta\tilde\Phi}\Big\rangle . $$ The free case is the statement that the expectation of the free field solves the classical equation, and the interaction contributes the source term on the right. The algebra's content is that the operator acting is the central $\Box-m^2$ of the series conventions, so that in momentum space the equation is the algebraic one $\mathcal M(\tilde k)\langle\tilde\Phi(k)\rangle=\cdots$ with $\mathcal M(\tilde k)=\tilde k\tilde k^{\natural}+m^2$ and $\tilde k=iE e_0+\mathbf p$ — the central mass-shell operator of The Feynman Propagator in Biquaternionic Form and of the functional-integral article. Nothing about the equation is new; what is new is the operator's address in the algebra, the center, and the module on which the expectation is taken.

The Gauge Field and the Fermion

The gauge field. The Schwinger–Dyson statement of the quantum Yang–Mills equation is derived in full in The Yang–Mills Path Integral and the Faddeev–Popov Procedure in Biquaternionic Form, and it is the same total-derivative argument with $\mathcal{O}[\mathcal{A}]$ in place of $F$: $$ \Big\langle\big(D_\mu F^{\mu\nu}\big)^a\Big\rangle=\Big\langle J^{\nu a}_{\mathrm{gf}}\Big\rangle+\Big\langle J^{\nu a}_{\mathrm{gh}}\Big\rangle , $$ with the gauge content carried by the Slavnov–Taylor identities rather than by the equation. The algebra's part is the covariant divergence as the adjoint twisting of a central differential operator; the two-sided structure of the gravitational analogue, and its consequences, belong to the quantum-gravity agenda and are not repeated here. The abelian limit of the same derivation is the Schwinger–Dyson statement of Maxwell's equation, with the field-independent Faddeev–Popov determinant of the abelian theory.

The fermion. For the Dirac field the quadratic operator is not central, because the mass term is the linear, chirality-off-diagonal right multiplication $$ \tilde\nabla\tilde\Psi_R=m\tilde\Psi_L,\qquad \tilde\nabla^{\natural}\tilde\Psi_L=m\tilde\Psi_R , $$ of Conventions in the Biquaternion Universe. The variation of the fermionic action therefore couples the two minimal left ideals, and the corresponding Schwinger–Dyson hierarchy mixes the chiralities: the equation of motion of $\tilde\Psi_L$ carries the mass times $\tilde\Psi_R$ and conversely, so the contact terms inherit the off-diagonal structure. This is the same non-central feature that makes the one-loop effective action mix the chiralities in The Generating Functional and the Effective Action in Biquaternionic Form, and it is the algebraic origin of the axial Ward identity's anomalous right-hand side below. The Grassmann-valued functional integral and its determinant are the subject of The Fermionic Path Integral in Biquaternionic Form and are used and not rebuilt.

The Ward–Takahashi Identities

The General Derivation

Let $Q$ be a derivation generating a one-parameter group of transformations of the field, $\delta\tilde\Phi=Q\tilde\Phi$, and suppose that the action is invariant in the local sense that $$ Q S=\int d^4x\;\partial_\mu j^\mu(x) $$ for some current $j^\mu$ — that is, $QS$ is a total divergence, so that the transformation is a symmetry of the bulk action up to a boundary term. Take $F$ in the working identity and replace the generic variation by the symmetry variation $Q$: $$ 0=\int\mathcal{D}\tilde\Phi\;Q\Big(F\,e^{\,iS/\hbar}\Big) =\int\mathcal{D}\tilde\Phi\;e^{\,iS/\hbar}\Big(QF+\frac{i}{\hbar}F\,QS\Big), $$ where the second equality uses the invariance of the measure, $\int\mathcal{D}\tilde\Phi\,Q(\;\cdot\;)=0$ — the assumption that the measure, and not only the action, is invariant. Then $$ \big\langle QF\big\rangle+\frac{i}{\hbar}\Big\langle F\int d^4x\,\partial_\mu j^\mu\Big\rangle=0 . $$ For $F=1$ this is the statement that the current is conserved inside the vacuum: $$ \Big\langle\partial_\mu j^\mu(x)\Big\rangle=0 . $$ For $F$ a product of fields the identity relates the divergence of the current to the variation of the product, and when $Q$ is a local generator, $Q=\int d^4x\,q(x)$, the identity holds locally as $$ \partial_\mu\big\langle j^\mu(x)\,F\big\rangle =-\big\langle q(x)F\big\rangle+\text{contact terms}, $$ the Ward–Takahashi identity. The contact terms are the variation of the inserted fields, exactly as in the Schwinger–Dyson case, and they are distributional at coincident points.

The premise is measure invariance. The step from the action's invariance to the identity is the invariance of the measure, and it is the step that can fail. Where the measure is not invariant, $QS$ need not vanish inside correlators and the divergence carries an extra term — the anomaly. The algebra does not supply the measure, so this is the premise the framework cannot check; what it supplies is the generator $Q$, the current's home, and the form of the symmetry.

The Central Phase

The framework's canonical one-parameter symmetry is the central phase, $$ \tilde\Phi\longmapsto e^{i\alpha}\tilde\Phi, \qquad Q\tilde\Phi=i\tilde\Phi , $$ generated by the central imaginary $i$, which is the algebra's continuous symmetry because $i$ is central and $e^{i\alpha}$ commutes with every element of $\mathbb{B}$. The Noether current of this symmetry was derived in Noether's Theorem in Biquaternionic Form: $$ \tilde J=i\Big[\tilde\Phi^{*}(\tilde\nabla\tilde\Phi)-(\tilde\nabla\tilde\Phi^{*})\tilde\Phi\Big]=J^\nu e_\nu\in\mathbb{M}_-, $$ with the Noether identity $\mathrm{Sc}(\tilde\nabla^{\natural}\tilde J)=i\,\mathrm{Sc}[\tilde\Phi^{*}(\Box\tilde\Phi)-(\Box\tilde\Phi^{*})\tilde\Phi]$, zero on shell. The Ward–Takahashi identity of the central phase is the quantum reading of that identity: $$ \partial_\mu\big\langle J^\mu(x)\,F\big\rangle =i\big\langle\tilde\Phi(x)F\big\rangle+\text{contact terms}, $$ the current divergence given by the variation of the inserted fields. The algebra supplies three things here and no more: the symmetry, because the center is $\mathbb{C}$ and its unit circle is a continuous group; the current's address, in the material sector $\mathbb{M}_-$; and the compactness of the identity, because the divergence is the scalar part of $\tilde\nabla^{\natural}\tilde J$. What it does not supply is the invariance of the measure over $\mathbb{B}$-valued fields, which is the premise of the identity and is transcribed.

The abelian Ward identity. In the gauge theory the central phase is the $U(1)$ of the abelian connection $\tilde A=ict\cdots$ of Maxwell's Equations in the Biquaternionic Formulation and Canonical Quantization of the Biquaternion Maxwell Field. The Ward identity of that symmetry is the statement that the longitudinal part of the photon's vacuum polarisation vanishes and that the vertex renormalisation equals the wave-function renormalisation, $Z_1=Z_2$; in the framework these are standard results of the abelian theory, transcribed, and their content is the conservation of the current that the algebra locates in $\mathbb{M}_-$. The non-abelian generalisation replaces the single identity by one per external leg and by the Slavnov–Taylor family, below.

The Axial Identity and the Anomaly

The second symmetry of the free massless Dirac field is the axial or chirality rotation, $\tilde\Psi\mapsto e^{i\alpha\gamma_5}\tilde\Psi$, generated by the chirality operator, whose conservation the mass breaks. The classical identity is $$ \partial_\mu j_5^\mu=2im\,\tilde\Psi^{\natural}\gamma_5\tilde\Psi , $$ holding for the linear, chirality-off-diagonal mass term of the series conventions, and it is recorded in Noether's Theorem in Biquaternionic Form as the divergence that the mass generates between the two central ideals of $\mathbb{B}$. The quantum identity is the Ward–Takahashi identity of the same rotation, and its right-hand side acquires the anomaly, $$ \partial_\mu\big\langle j_5^\mu\big\rangle=2im\big\langle\tilde\Psi^{\natural}\gamma_5\tilde\Psi\big\rangle+\frac{g^2}{16\pi^2}\big\langle F\tilde F\big\rangle , $$ the second term being the topological density. In the quantum theory the axial current is not conserved even at $m=0$, because the measure of the fermionic functional integral is not invariant under the chirality rotation — precisely the failure of the measure premise. The biquaternion reading is the one of Anomalies and Anomaly Cancellation in Biquaternionic Form and The Fujikawa Method and the Path-Integral Anomaly in Biquaternionic Form: the anomaly is the non-invariance of the fermionic measure, computed from the Jacobian of the rotation on the module, and the algebra's part is the module's chirality structure and the trace that defines the Jacobian. The algebra does not produce the anomaly's coefficient; it locates the object whose Jacobian carries it.

Non-Abelian: the Slavnov–Taylor Identities

For a non-abelian gauge theory the symmetry whose Ward–Takahashi identity is at issue is the BRST symmetry of BRST Symmetry in Biquaternionic Form, and its identities are the Slavnov–Taylor identities. They are one identity for each external leg, relating amplitudes with different numbers of ghosts and longitudinal gauge fields, and they express the gauge-parameter independence of physical amplitudes and the decoupling of longitudinal modes. The framework's contribution to them is exactly its contribution to BRST: the gauge algebra realized by the commutator, the Jacobi identity inherited from associativity, and hence the nilpotency $s^2=0$ that makes the identities close. The identities themselves, their renormalisation, and the four-dimensional consistency conditions are standard and are transcribed; they are recorded here as the non-abelian face of the same measure-invariance argument, with $Q$ the odd derivation $s$.

What the Algebra Supplies, What It Imports, What It Does Not Supply

Supplied by the algebra, and recomputed here. The functional derivative as the module gradient defined through the real bilinear form, so that the total-derivative argument needs no basis and no coordinate chart; the variation of the quadratic action as the central operator $\tilde K=\Box-m^2$ acting on the module, so that the Schwinger–Dyson hierarchy is generated by the central d'Alembertian and the mass; the chirality-off-diagonal form of the fermionic variation, from the right-multiplication mass term; the central phase as the algebra's canonical one-parameter symmetry, with the generator $i$ and the current in $\mathbb{M}_-$; the abelian Ward identity as its quantum reading; and the module's chirality structure as the carrier of the axial anomaly's Jacobian. The finite-dimensional instance of the master equation was checked with an explicit Hermitian matrix, and the free propagator was recovered with the sign recorded above.

Imported, and left visible. The total-derivative argument itself; the assumption of translation invariance of the functional measure, without which neither family of identities holds; the contact terms and their distributional character; the master equation and the $J$-expansion of the hierarchy; the local form of the Ward–Takahashi derivation; the abelian Ward identity $Z_1=Z_2$ and the vanishing of the longitudinal vacuum polarisation; the axial anomaly and its coefficient; the Slavnov–Taylor identities and their renormalisation. These are standard field theory (Schwinger, Dyson, Ward, Takahashi, Zinn-Justin, Peskin–Schroeder, Weinberg) and are not rebuilt.

Not supplied. The functional measure and its invariance under translation or under a symmetry; the space of field configurations; the regulator; and any empirical content. The framework's identities are conditional on the measure premise, which is exactly the premise the algebra cannot formulate, and the article states the condition rather than closing it. The conditional character is not a defect of the derivation — the same condition holds in the standard theory and is the reason anomalies exist there — but it is the honest boundary of what the algebra can check.

Open Questions

A measure the algebra can formulate. The identities rest on the invariance of $\mathcal{D}\tilde\Phi$, and the algebra carries no measure. Is there a finitely generated or algebraic measure on $\mathbb{B}$-valued configurations — a Gaussian weight fixed by the trace form and the central kinetic operator — whose translation invariance is algebraic rather than analytic, and does it reproduce the identities of this article without the analytic premise?
The contact terms and the sector split. The contact terms are delta functions at coincident points, and the fields carry the sector split $\mathbb{M}_-\oplus\mathbb{M}_+$. Does the split organise the contact terms into two families — a sector-diagonal one for the central kinetic operator and a sector-mixing one for the Dirac mass — and does the mixing family account for the axial anomaly's structure?
The master equation on the module. The master equation was written with the variation evaluated at $\delta/\delta\tilde J$ through the real pairing. Can the whole hierarchy be packaged as a single identity of $\mathbb{B}$-module operators, without passing through a component basis, in the way the effective action is packaged on the module?
The gravitational Ward identity. The gauge Ward identity comes from a one-sided symmetry of a one-sided connection; the gravitational analogue has the two-sided rotor action of Curved Spacetime and the Biquaternion Framework. What is the Ward–Takahashi identity of the local Lorentz symmetry, and does its derivation meet the same two-sided obstruction that the quantum-gravity agenda records for the gauge-fixing problem?
The central phase and the axial identity as a pair. Both identities are Ward–Takahashi identities of an internal rotation, one of the center and one of the chirality. The center's identity is exact; the chirality's acquires the anomaly. Is the difference the algebra's, in the sense that the center is a genuine scalar symmetry and the chirality is the off-diagonal rotation between the two central ideals, or is it the standard difference of measure behaviour that the algebra merely inherits?
Empirical content. As everywhere in the framework, the identities are the standard ones in new notation. Whether the biquaternion packaging imposes any constraint that the four-vector formulation does not is not established here.

Summary

The Schwinger–Dyson equations and the Ward–Takahashi identities descend from a single argument, the integration of a total derivative in field space, and both rest on the translation invariance of the functional measure. The working identity is $$ \Big\langle\frac{\delta F}{\delta\tilde\Phi(x)}\Big\rangle_{\!J}+\frac{i}{\hbar}\Big\langle F\,\frac{\delta S}{\delta\tilde\Phi(x)}\Big\rangle_{\!J}+\big\langle F\,\tilde J(x)\big\rangle_{\!J}=0, $$ which at $F=1$ gives the quantum equation of motion $\langle\delta S/\delta\tilde\Phi(x)\rangle=0$, at generic $F$ the contact-term hierarchy, and at the symmetry variation gives the Ward–Takahashi identity $\partial_\mu\langle j^\mu F\rangle=-\langle qF\rangle+\text{contact terms}$. The generating form is the master Schwinger–Dyson equation $$ \Big(\frac{\delta S}{\delta\tilde\Phi}\Big)\Big[\frac{\delta}{\delta\tilde J}\Big]Z[\tilde J]=i\hbar\,\tilde J(x)\,Z[\tilde J], $$ whose sign is fixed by the series convention $Z=\int\mathcal{D}\tilde\Phi\,e^{\,iS/\hbar+\langle\tilde J,\tilde\Phi\rangle}$ with a real source pairing; its $J$-expansion generates the whole hierarchy, and the free case was checked in finite dimension.

The algebra's contribution is specific and limited. It supplies the functional derivative as the module's own gradient, so that the argument needs no basis; the variation of the quadratic action as the central $\tilde K=\Box-m^2$ on the module, so that the scalar hierarchy is generated by the central d'Alembertian and the mass; the chirality-off-diagonal fermionic variation, from the right-multiplication mass term, which is the algebraic origin of the axial anomaly's structure; the central phase as the canonical one-parameter symmetry, with generator $i$ and current in the material sector $\mathbb{M}_-$; and the module's chirality structure as the carrier of the anomaly's Jacobian. It does not supply the functional measure or its invariance, which is the premise of every identity here; the contact terms; the master equation; the local Ward–Takahashi derivation; the anomaly coefficient; or the Slavnov–Taylor identities of the non-abelian theory. The quantum Yang–Mills equation of the companion article is the gauge-field instance of the same argument, the axial anomaly of the anomaly articles is the failure of its measure premise, and the BRST nilpotency of the BRST article is what makes the non-abelian identities close.

Summary of Notation

Symbol	Meaning
$\mathbb{B}=\mathbb{C}\otimes_\mathbb{R}\mathbb{H}$	Biquaternion algebra, $\cong M_2(\mathbb{C})$
$e_0=1,e_1,e_2,e_3$	Quaternion basis, $e_k^2=-e_0$
$i$	Central scalar imaginary, $i^2=-1$
$\mathbb{M}_-,\mathbb{M}_+$	Material (anti-Hermitian) and informational (Hermitian) sectors
$\flat=-{}^{*}$	Anti-Hermitian conjugation
$\langle\tilde Q,\tilde Y\rangle=\mathrm{Re}\,\mathrm{Tr}(\tilde Q^{*}\tilde Y)$	Real bilinear form; source pairing
$\mathrm{Tr}(\tilde P\tilde H)=2\,\mathrm{Sc}(\tilde P\tilde H)$, $\mathrm{Tr}(e_0)=2$	Trace formula
$\Box=\tilde\nabla\tilde\nabla^{\natural}=\partial_{ict}^2+\Delta$	d'Alembertian, series convention
$\tilde K=\Box-m^2$	Quadratic operator
$Z[\tilde J]=\int\mathcal{D}\tilde\Phi\,e^{\,iS/\hbar+\langle\tilde J,\tilde\Phi\rangle}$	Functional integral with module-valued source
$\delta F/\delta\tilde\Phi$	Module gradient, defined through the pairing
$\langle\delta S/\delta\tilde\Phi(x)\rangle=0$	Quantum equation of motion
$\big(\delta S/\delta\tilde\Phi\big)[\delta/\delta\tilde J]Z=i\hbar\tilde JZ$	Master Schwinger–Dyson equation
$Q$, $Q\tilde\Phi=i\tilde\Phi$	Symmetry generator; central phase generator
$j^\mu$, $\tilde J=J^\nu e_\nu\in\mathbb{M}_-$	Noether current of the central phase
$\partial_\mu\langle j^\mu F\rangle=-\langle qF\rangle+\text{contact terms}$	Ward–Takahashi identity
$\partial_\mu j_5^\mu=2im\,\tilde\Psi^{\natural}\gamma_5\tilde\Psi$	Classical axial divergence
$g^2\langle F\tilde F\rangle/16\pi^2$	Axial anomaly (topological density)
$s$, $s^2=0$	BRST derivation and nilpotency; source of the Slavnov–Taylor identities
$\mathcal{D}\tilde\Phi$	Functional measure; its invariance is the premise