Superspace as a way to organize known fields

The free chiral action and its superpotential interaction contain several terms whose coefficients supersymmetry ties together. A superfield puts those component fields into one polynomial in anticommuting coordinates. Multiplying and integrating these polynomials will recover the actions we already checked.

This is a flat, three-dimensional Euclidean calculation with the complexified $\mathcal N=2$ algebra. Its Lorentzian parent has four real supercharges. There are no fixed background fields or central charge here. The first calculations use a gauge-neutral chiral multiplet; an Abelian gauge superfield enters near the end. Superspace will not impose a Euclidean integration contour or add physical dimensions to spacetime.

Continue from the vector multiplet. We reuse the ordered integration in Grassmann calculus, the matrices in three-dimensional spinors, and the component results of the free theory and superpotential interactions. No four-dimensional superspace is assumed.

A polynomial that remembers an interaction

Introduce two independent odd spinor coordinates, $\theta_\alpha$ and $\widetilde\theta_\alpha$, with $\alpha=1,2$. They anticommute with each other and with the component fermions. A bosonic superfield is even: a term such as $\theta\psi$ contains two odd factors. Its coefficients remain fields on ordinary three-dimensional spacetime.

Our spinor contraction is $uv=u^\alpha v_\alpha=u^TCv$, where

$$C=\begin{pmatrix}0&-1\\1&0\end{pmatrix}, \qquad u^\alpha=E^{\alpha\beta}u_\beta, \qquad E=\begin{pmatrix}0&1\\-1&0\end{pmatrix}.$$

Consequently,

$$\theta^2\equiv\theta\theta=-2\theta_1\theta_2, \qquad \widetilde\theta^2=-2\widetilde\theta_1\widetilde\theta_2.$$

Keep the elementary orientation $\int d\theta_2\,d\theta_1\,\theta_1\theta_2=1$. To give the compact spinor measure the target paper’s normalization, define

$$\int d^2\theta=-\frac12\int d\theta_2\,d\theta_1, \qquad \int d^2\widetilde\theta=-\frac12 \int d\widetilde\theta_2\,d\widetilde\theta_1.$$

Thus $\int d^2\theta\,\theta^2=\int d^2\widetilde\theta\,\widetilde\theta^2=1$. Also define

$$\int d^4\theta =\frac14\int d\widetilde\theta_2\,d\widetilde\theta_1\, d\theta_2\,d\theta_1, \qquad \int d^4\theta\,\theta^2\widetilde\theta^2=1.$$

The factor $1/4$ and the elementary ordering are both part of this definition. These are the three-manifold paper v3, (A.6), printed p. 38 expressed in our explicit ordered basis.

For the moment set $\widetilde\theta=0$. The chiral polynomial packages the familiar fields as

$$\Phi\big|_{\widetilde\theta=0} =\phi+\sqrt2\,\theta\psi+\theta^2F_\Phi.$$

Here $F_\Phi$ denotes the auxiliary field, to distinguish it from a superfield or an integral. We will derive the missing $\widetilde\theta$ dependence shortly. Start with the quadratic superpotential $W(\Phi)=m\Phi^2/2$. Only three products contribute to its $\theta^2$ coefficient:

$$\frac m2\Phi^2\big|_{\theta^2} =m\phi\,\theta^2F_\Phi +m(\theta\psi)^2.$$

The sign of the second term follows from actual odd reordering:

$$\theta\psi=\theta_2\psi_1-\theta_1\psi_2, \qquad (\theta\psi)^2=-2\theta_1\theta_2\psi_1\psi_2 =-\frac12\theta^2\psi\psi.$$

Therefore

$$\boxed{\int d^2\theta\,\frac m2\Phi^2 =m\phi F_\Phi-\frac m2\psi\psi.}$$

The integral is evaluated on the chiral slice $\widetilde\theta=0$, or equivalently at fixed chiral coordinate defined below. It has recovered the auxiliary coupling and Yukawa term of the component mass interaction. Treating the fermion coefficients as commuting numbers would give the wrong result.

For an analytic $W$, expand around the even bottom component $\phi$. The nilpotent increment is $\sqrt2\,\theta\psi+\theta^2F_\Phi$; only its first and second powers contribute. Hence

$$\int d^2\theta\,W(\Phi) =W'(\phi)F_\Phi-\frac12W''(\phi)\psi\psi.$$

For the previous model, $W=m\phi^2/2+g\phi^3/3$, this becomes

$$(m\phi+g\phi^2)F_\Phi -\frac12(m+2g\phi)\psi\psi.$$

Superspace has packaged a known relation between coefficients. It has not changed the scalar potential or replaced the auxiliary-field calculation.

What makes the polynomial chiral

Chirality is a differential constraint on the polynomial. To state it, keep the matrices

$$\gamma^1=\begin{pmatrix}1&0\\0&-1\end{pmatrix}, \quad \gamma^2=\begin{pmatrix}0&-1\\-1&0\end{pmatrix}, \quad \gamma^3=\begin{pmatrix}0&i\\-i&0\end{pmatrix}.$$

They obey $\gamma^\mu\gamma^\nu=\delta^{\mu\nu}+i\epsilon^{\mu\nu\rho}\gamma_\rho$, with $\epsilon^{123}=1$. All odd derivatives below are left derivatives. Define the supercovariant derivatives

$$D_\alpha=\frac{\partial^L}{\partial\theta^\alpha} -i(\gamma^\mu\widetilde\theta)_\alpha\partial_\mu, \qquad \widetilde D_\alpha=-\frac{\partial^L}{\partial\widetilde\theta^\alpha} +i(\gamma^\mu\theta)_\alpha\partial_\mu.$$

The odd derivative acts on a raised coordinate. In the independent lower-coordinate basis this means

$$\frac{\partial^L}{\partial\theta^1} =\frac{\partial^L}{\partial\theta_2}, \qquad \frac{\partial^L}{\partial\theta^2} =-\frac{\partial^L}{\partial\theta_1},$$

and likewise for $\widetilde\theta$. This distinction is necessary to reproduce the three-manifold paper v3, (A.5).

These operators are odd. Their graded Leibniz rule is the one used for Grassmann differentiation. Their algebra is

$$\{D_\alpha,D_\beta\}=0, \qquad \{\widetilde D_\alpha,\widetilde D_\beta\}=0, \qquad \{D_\alpha,\widetilde D_\beta\} =-2i(\gamma^\mu)_{\alpha\beta}\partial_\mu,$$

where the matrix with two lower indices is $\gamma^\mu E$. For example, the constant term in the mixed anticommutator comes from differentiating the odd coordinate in the other operator; the terms with two spacetime derivatives cancel by odd anticommutation. These $D$ operators test constraints; they are not additional supercharges or the ordinary matter covariant derivative.

Define the even nilpotent vector

$$b^\mu=\theta\gamma^\mu\widetilde\theta, \qquad y^\mu=x^\mu-i b^\mu, \qquad \widetilde y^\mu=x^\mu+i b^\mu.$$

Direct differentiation gives $\widetilde D_\alpha y^\mu=0$ and $D_\alpha\widetilde y^\mu=0$. For instance,

$$b^1=\theta_1\widetilde\theta_2+\theta_2\widetilde\theta_1, \qquad \widetilde D_1y^1 =-\partial^L_{\widetilde\theta_2}(x^1-i b^1)+i\theta_1=0.$$

The derivative $\partial^L_{\widetilde\theta_2}(\theta_1\widetilde\theta_2)=-\theta_1$ supplies the essential sign. A chiral superfield satisfies $\widetilde D_\alpha\Phi=0$. In the coordinates $(y,\theta,\widetilde\theta)$ the constraint removes independent $\widetilde\theta$ dependence, giving

$$\Phi(y,\theta)=\phi(y)+\sqrt2\,\theta\psi(y)+\theta^2F_\Phi(y).$$

This explains why there are only the familiar three component fields. An antichiral superfield satisfies $D_\alpha\widetilde\Phi=0$. To match the previously fixed component transformations, its expansion is

$$\widetilde\Phi(\widetilde y,\widetilde\theta) =\widetilde\phi(\widetilde y) -\sqrt2\,\widetilde\theta\widetilde\psi(\widetilde y) -\widetilde\theta^2\widetilde F_\Phi(\widetilde y).$$

Both minus signs follow from the embeddings in the three-manifold paper v3, (6.7) and (6.9), together with its flat expansion (6.4). The tilded coefficients are independent Euclidean variables. A tilde does not impose pointwise complex conjugation.

The engineering dimensions are $[\theta]=[\widetilde\theta]=-1/2$. Then $[\phi]=1/2$, $[\psi]=1$, and $[F_\Phi]=3/2$ make every term in $\Phi$ have dimension $1/2$. The measure dimensions are $[d^2\theta]=1$ and $[d^4\theta]=2$. Thus a superpotential has dimension two and $\widetilde\Phi\Phi$ has dimension one, as required for a three-dimensional Lagrangian of dimension three.

If an R-symmetry is present, assign $R(\theta)=1$ and $R(\widetilde\theta)=-1$. A superfield of charge $r$ then has component charges $r,r-1,r-2$. Its chiral integral is R-invariant when $W$ has charge two. As in the component lesson, generic fixed nonzero quadratic and cubic couplings need not preserve a continuous R-symmetry. Flat supersymmetry still holds; no R-symmetric curved completion is implied.

Expand at ordinary spacetime coordinates

A product $\widetilde\Phi\Phi$ involves both $y$ and $\widetilde y$. To integrate it, first express both factors at the same $x$. The useful finite identities are

$$b^\mu b^\nu=\frac12\delta^{\mu\nu}\theta^2\widetilde\theta^2, \qquad b^\mu(\theta\psi)=\frac12\theta^2 (\widetilde\theta\gamma^\mu\psi).$$

The first identity can be checked one matrix at a time: $(b^1)^2=2\theta_1\theta_2\widetilde\theta_1\widetilde\theta_2$, while $b^1b^2=0$. The other directions give the same diagonal coefficient. Taylor expansion therefore stops after two spacetime derivatives:

$$\phi(x-i b)=\phi-i b^\mu\partial_\mu\phi -\frac14\theta^2\widetilde\theta^2\partial^2\phi, \qquad \partial^2=\delta^{\mu\nu}\partial_\mu\partial_\nu.$$

The shift of the spinor contributes the cubic term; the shift of $F_\Phi$ vanishes after multiplication by $\theta^2$. The complete result is

$$\begin{aligned} \Phi={}&\phi+\sqrt2\,\theta\psi+\theta^2F_\Phi -i b^\mu\partial_\mu\phi\\ &-\frac{i}{\sqrt2}\theta^2 \widetilde\theta\gamma^\mu\partial_\mu\psi -\frac14\theta^2\widetilde\theta^2\partial^2\phi,\\[2pt] \widetilde\Phi={}&\widetilde\phi -\sqrt2\,\widetilde\theta\widetilde\psi -\widetilde\theta^2\widetilde F_\Phi +i b^\mu\partial_\mu\widetilde\phi\\ &+\frac{i}{\sqrt2}\widetilde\theta^2 \theta\gamma^\mu\partial_\mu\widetilde\psi -\frac14\theta^2\widetilde\theta^2\partial^2\widetilde\phi. \end{aligned}$$

This is an exact nilpotent expansion, not a slow-variation approximation. Arbitrarily rapid spacetime dependence is allowed wherever the component fields are smooth.

Now select the coefficient of $\theta^2\widetilde\theta^2$ in $\widetilde\Phi\Phi$. Each contribution has a definite origin:

Products being multiplied	Contribution to the integrated coefficient
The two scalar bilinear shifts	$\frac12\,\partial_\mu\widetilde\phi\,\partial^\mu\phi$
A bottom scalar and the other scalar’s quartic shift	$-\frac14\widetilde\phi\,\partial^2\phi-\frac14\phi\,\partial^2\widetilde\phi$
A linear fermion and the other fermion’s cubic shift	$-\frac i2\widetilde\psi\gamma^\mu\partial_\mu\psi+\frac i2(\partial_\mu\widetilde\psi)\gamma^\mu\psi$
The two auxiliary terms	$-\widetilde F_\Phi F_\Phi$

For example, the last sign is already visible in $(-\widetilde\theta^2\widetilde F_\Phi)(\theta^2F_\Phi)$; commuting the two even quadratic monomials introduces no further sign.

Use the ordinary product rule on the scalar and fermion expressions in the table. The full answer is

$$\begin{aligned} \int d^4\theta\,\widetilde\Phi\Phi ={}&\partial_\mu\widetilde\phi\,\partial^\mu\phi -i\widetilde\psi\gamma^\mu\partial_\mu\psi -\widetilde F_\Phi F_\Phi+\partial_\mu B^\mu,\\ B^\mu={}&\frac i2\widetilde\psi\gamma^\mu\psi -\frac14\partial^\mu(\widetilde\phi\phi). \end{aligned}$$

The first three terms are exactly the Lagrangian derived in the free-chiral lesson. The equality of actions requires the integral of $\partial_\mu B^\mu$ to vanish, for example for periodic fields or suitable falloff. On a space with a boundary one must keep the displayed surface term and specify boundary conditions.

The familiar auxiliary sign remains negative. The free convergent Gaussian may use the component contour $\widetilde F_\Phi=-F_\Phi^*$; with interactions the shifted contour described in the superpotential lesson applies. Superspace notation does not choose either contour.

For the antichiral superpotential the same Taylor calculation gives

$$\int d^2\widetilde\theta\,\widetilde W(\widetilde\Phi) =-\widetilde W'(\widetilde\phi)\widetilde F_\Phi -\frac12\widetilde W''(\widetilde\phi) \widetilde\psi\widetilde\psi.$$

Thus the flat interacting action in our component convention is

$$S=\int d^3x\left[ \int d^4\theta\,\widetilde\Phi\Phi +\int d^2\theta\,W(\Phi) -\int d^2\widetilde\theta\,\widetilde W(\widetilde\Phi) \right],$$

up to the specified surface term. The minus sign before the antichiral integral produces the plus $\widetilde F_W$ in the component formula $F_W+\widetilde F_W$ of the three-manifold paper v3, (6.14). It is a convention translation, not a negative physical scalar potential.

Why are these integrals useful beyond this example? Flat supertranslations act by odd-coordinate derivatives and spacetime derivatives. A full Berezin integral annihilates the former, because differentiation removes one of the variables needed for the top coefficient. The remaining terms are spacetime divergences. On a chiral slice the same reasoning applies to its two independent odd coordinates. This is the integration argument in Bilal v1, §4.3, (4.25)–(4.29); its four-dimensional matrices and component signs are not being imported here. Our explicit three-dimensional expansion fixes the coefficients, and the free and interacting chiral lessons supply their component invariance checks.

General and linear packages

A general superfield is an unconstrained polynomial in all four odd coordinates. There are $1+6+1=8$ even monomials and $4+4=8$ odd monomials. For an even superfield their coefficient functions are respectively bosonic and fermionic. With complex coefficients this is $8+8$ complex components before any constraints. It is a count of coefficient functions, not of propagating particle modes.

Write its lowest coefficient as $C_{\mathcal S}$ and its $\theta^2\widetilde\theta^2$ coefficient as $T_{\mathcal S}$. The target paper names the improved highest component $D_{\mathcal S}$ by

$$T_{\mathcal S} =-\frac12\left(D_{\mathcal S} +\frac12\partial^2 C_{\mathcal S}\right).$$

Consequently, $\int d^4\theta\,\mathcal S$ gives $-D_{\mathcal S}/2$ up to a divergence. This translates the flat expansion the three-manifold paper v3, (6.4) into its D-term formula (6.13). A D-term integral is a prescription for selecting a component; its name does not mean that the superfield must be a gauge multiplet.

A linear superfield imposes two second-order odd constraints. We use the unambiguous component notation

$$D_1D_2\mathcal J=0, \qquad \widetilde D_1\widetilde D_2\mathcal J=0.$$

The rightmost operator acts first. The usual contracted notation $D^2\mathcal J=0$ differs only by a nonzero constant factor. We take $\mathcal J$ to have zero R-charge and zero central charge. Its independent components are two scalars $J,K$, two odd spinors $j,\widetilde j$, and a conserved vector $j_\mu$. In the flat specialization of the three-manifold paper v3, (6.11), the expansion is

$$\begin{aligned} \mathcal J={}&J+i\theta j+i\widetilde\theta\widetilde j -b^\mu j_\mu+i\theta\widetilde\theta K\\ &-\frac12\widetilde\theta^2\theta\gamma^\mu\partial_\mu\widetilde j -\frac12\theta^2\widetilde\theta\gamma^\mu\partial_\mu j +\frac14\theta^2\widetilde\theta^2\partial^2J. \end{aligned}$$

The cubic and quartic coefficients are fixed derivatives of lower components. They are not extra fields to integrate independently. To see what the constraints impose, temporarily allow an arbitrary $j_\mu$. Applying the explicitly defined odd derivatives gives

$$D_1D_2\mathcal J =i\widetilde\theta^2\,\partial_\mu j^\mu, \qquad \widetilde D_1\widetilde D_2\mathcal J =-i\theta^2\,\partial_\mu j^\mu.$$

All other coefficients cancel. Thus both constraints hold precisely when $\partial_\mu j^\mu=0$. The name “linear” refers to these linear differential constraints; $\mathcal J$ may itself be a composite of interacting fields.

This package can contain a conserved flavor current. If it is built from matter fields, the conservation law may use those fields’ equations of motion. A different example below obeys the same law identically. The source calls this a real linear multiplet, reflecting its Lorentzian reality structure. We keep its Euclidean coefficients complex until a particular reality condition or contour is stated.

The top coefficient also gives a useful warning:

$$\int d^4\theta\,\mathcal J=\frac14\partial^2J.$$

Integrating a lone linear superfield over all of superspace produces a boundary term. It does not supply a kinetic action merely because the package contains several fields. Nor is this scalar linear multiplet the full multiplet containing the stress tensor and supersymmetry current.

The vector multiplet in this language

The Abelian vector multiplet introduced in the component lesson fits into a general superfield $\mathcal V$ with a gauge redundancy. In Wess–Zumino gauge, the removable lower components have been set to zero, leaving

$$\mathcal V=b^\mu a_\mu-i\theta\widetilde\theta\,\sigma +i\widetilde\theta^2\theta\lambda -i\theta^2\widetilde\theta\widetilde\lambda -\frac12\theta^2\widetilde\theta^2D_{\rm v}.$$

Here $a_\mu$ is the dynamical Abelian connection, $\sigma$ the real scalar on the usual bosonic contour, and $D_{\rm v}$ the auxiliary field. The subscript separates this field from the derivative $D_\alpha$. It is the same auxiliary called $D$ in the vector-multiplet lesson and the paper. The superspace gauge freedom before fixing this gauge is

$$\mathcal V\longmapsto\mathcal V+\Omega+\widetilde\Omega, \qquad \widetilde D_\alpha\Omega=0, \quad D_\alpha\widetilde\Omega=0.$$

For matter of charges $q$ and $-q$, respectively,

$$\Phi\longmapsto e^{2q\Omega}\Phi, \qquad \widetilde\Phi\longmapsto e^{2q\widetilde\Omega}\widetilde\Phi.$$

Since these are Abelian even superfields, their exponentials commute. The product

$$\widetilde\Phi\,e^{-2q\mathcal V}\Phi$$

is gauge invariant: the exponent changes cancel exactly. This packages the gauge-covariant matter coupling. For the residual ordinary transformation, $\Omega|=i\alpha/2$ and $\widetilde\Omega|=-i\alpha/2$, so its bottom components transform as $\phi\mapsto e^{iq\alpha}\phi$ and $\widetilde\phi\mapsto e^{-iq\alpha}\widetilde\phi$, while $a_\mu\mapsto a_\mu+\partial_\mu\alpha$. These are the conventions of the three-manifold paper v3, (6.17)–(6.22). Supersymmetry generally requires a compensating gauge transformation to retain Wess–Zumino gauge, as the component closure in the vector-multiplet lesson anticipated.

The gauge-invariant field-strength superfield $\Sigma$ is linear. Its components are the preceding linear package with

$$J=\sigma,\qquad j=i\widetilde\lambda,\qquad \widetilde j=-i\lambda,\qquad j^\mu=-\frac i2\epsilon^{\mu\nu\rho}f_{\nu\rho}, \qquad K=D_{\rm v}.$$

This is the flat limit of the three-manifold paper v3, (6.21). Its conservation equation follows from the Abelian Bianchi identity:

$$\partial_\mu j^\mu =-i\epsilon^{\mu\nu\rho}\partial_\mu\partial_\nu a_\rho=0.$$

No Maxwell equation was used. The same superspace constraint can therefore describe a physical conserved current or an identically conserved dual field strength, with different reasons for conservation. Products of these packages will construct actions; choosing the package by itself does not choose Yang–Mills or Chern–Simons dynamics.

Exercises

1. Diagnose a measure error

A calculation replaces $d^2\theta$ by the unweighted ordered measure $d\theta_2\,d\theta_1$ while retaining $\theta^2=\theta\theta$. What does it obtain for the quadratic superpotential? Find the error and verify the dimensions of the correct answer.

Hint 1

First integrate $\theta^2$ itself. Its coefficient in the ordered basis is not one.

Hint 2

Both terms in the $\theta^2$ coefficient receive the same erroneous factor. A changed integration normalization is not a change to the Yukawa-to-auxiliary ratio.

Solution

Since $\theta^2=-2\theta_1\theta_2$, the unweighted measure gives $\int d\theta_2\,d\theta_1\,\theta^2=-2$. It therefore returns

$$-2m\phi F_\Phi+m\psi\psi =-2\left(m\phi F_\Phi-\frac m2\psi\psi\right).$$

Multiplying the measure by $-1/2$ restores the specified result. A different measure convention could be used consistently, but every action coefficient would have to be translated with it.

The chiral integral adds dimension one to $[W]=2$. With $[m]=1$, both $m\phi F_\Phi$ and $m\psi\psi$ have dimension three. For the cubic interaction $[g]=1/2$, so $g\phi^2F_\Phi$ and $g\phi\psi\psi$ also have dimension three. The normalization error passes this dimensional test; the ordered integral detects it.

2. Keep the independent antichiral signs

Take $W=m\Phi^2/2$ and $\widetilde W=\widetilde m\,\widetilde\Phi^2/2$. Derive both half-superspace integrals, combine them with the auxiliary term from the full integral, and eliminate $F_\Phi,\widetilde F_\Phi$. Then impose $\widetilde m=m^*$ and $\widetilde\phi=\phi^*$ on the scalar contour. What scalar mass term results?

Hint 1

The antichiral polynomial contains $-\widetilde\theta^2\widetilde F_\Phi$. Its linear fermion term has a minus sign too, but this one is squared in the Taylor expansion.

Hint 2

The action subtracts the antichiral integral. Vary the two auxiliary fields independently before imposing any contour.

Solution

The two integrals are

$$I_W=m\phi F_\Phi-\frac m2\psi\psi, \qquad I_{\widetilde W} =-\widetilde m\widetilde\phi\widetilde F_\Phi -\frac{\widetilde m}{2}\widetilde\psi\widetilde\psi.$$

The action uses $I_W-I_{\widetilde W}$. Its algebraic auxiliary sector is

$$\mathcal L_F=-\widetilde F_\Phi F_\Phi +m\phi F_\Phi+\widetilde m\widetilde\phi\widetilde F_\Phi.$$

Independent variation gives $\widetilde F_\Phi=m\phi$ and $F_\Phi=\widetilde m\widetilde\phi$. Substitution yields

$$\mathcal L_F\big|_{\rm stationary} =m\widetilde m\,\widetilde\phi\phi =|m|^2|\phi|^2$$

on the stated scalar contour. The Yukawa terms are $-m\psi\psi/2+\widetilde m\widetilde\psi\widetilde\psi/2$, agreeing with the component superpotential action. Algebraic elimination identifies the stationary action; the auxiliary Gaussian still needs the convergent shifted contour discussed there.

Using a plus between the two half-superspace integrals while holding all component definitions fixed would reverse the antichiral contribution and give the wrong stationary scalar sign. Neither a tilde nor a remembered “plus Hermitian conjugate” can replace the explicit convention translation.

3. Distinguish conservation from dynamics

In the proposed linear expansion set $J=K=j=\widetilde j=0$ and choose the vector profile $j_\mu=(x^1,0,0)$. Does it define a linear superfield? Contrast it with the field-strength assignment from a smooth Abelian connection. Finally explain why $\int d^3x\,d^4\theta\,\Sigma$ cannot be the Maxwell kinetic action under periodic or decaying boundary conditions.

Hint 1

Evaluate $\partial_\mu j^\mu$ before applying an equation of motion. The constraints can reject a proposed polynomial.

Hint 2

For the field-strength assignment, antisymmetry meets two commuting ordinary derivatives. For its full integral, use the lowest component $J=\sigma$.

Solution

The proposed vector has divergence one. Hence

$$D_1D_2\mathcal J=i\widetilde\theta^2, \qquad \widetilde D_1\widetilde D_2\mathcal J=-i\theta^2.$$

It is a general polynomial of the displayed form but fails the linear constraints. The constraints require conservation; the name of a coefficient does not guarantee it.

By contrast, $j^\mu=-i\epsilon^{\mu\nu\rho}\partial_\nu a_\rho$ is identically conserved for a smooth local connection. Commuting $\partial_\mu$ and $\partial_\nu$ makes their symmetric product vanish against $\epsilon^{\mu\nu\rho}$. This statement does not require the gauge field to satisfy its dynamical equations.

For $\Sigma$ the full integral is $\partial^2\sigma/4$. Its spacetime integral vanishes under the stated boundary conditions, even for a nonzero field strength with positive Maxwell action. A nontrivial gauge kinetic action must use an appropriate composite and normalization, rather than the integral of $\Sigma$ alone.

We can now read a superspace action by asking three concrete questions: which polynomial is being integrated, which ordered coefficient the measure selects, and which component convention fixes its sign. The next Abelian gauge-theory construction combines the vector package with a specified charged pair. As an immediate checkpoint, recover the invariant product $\widetilde\Phi e^{-2q\mathcal V}\Phi$ and explain why its superspace gauge invariance must reproduce the covariant derivative $\partial_\mu-iq a_\mu$ of the vector lesson.

Paper guides

Connect this lesson to the original papers and their notation.

• Paper guide: Section 2, currents and rigid backgrounds
• Paper guide: Appendices A and B, conventions and geometry