Consistent Truncations and Lower-Dimensional Effective Actions

A compactification truncation is consistent precisely when every solution of its retained lower-dimensional equations reconstructs a solution of the stated higher-dimensional equations. The decisive test is therefore nonlinear closure: after inserting the complete uplift ansatz, every discarded field equation, constraint, and Bianchi identity must vanish on the lower-dimensional solution. A large Kaluza–Klein mass does not prove this. It can instead justify a Wilsonian approximation in which heavy modes are solved for and their effects are retained with an error estimate.

Required background. Flux quantization and harmonic spectra supplies the Kaluza–Klein modes and compact-space data being tested. Two- and four-derivative supergravity EFT supplies the derivative-order and cutoff bookkeeping for the parent equations.

Helpful background. The ordered map across multiple thresholds develops the Wilsonian alternative. Validation and breakdown diagnostics explains how an approximate omission acquires a quantitative error model.

The nonlinear closure criterion

Let $x$ denote the $d$ noncompact coordinates, $y$ the coordinates on a compact space $K$, and $\Phi^A(x,y)$ all fields of a specified $D$-dimensional parent theory. The label $A$ includes tensor components and species. A truncation proposes an ansatz

$$\Phi^A(x,y)=\mathcal U^A[\phi^i](x,y)$$

in terms of finitely many lower-dimensional fields $\phi^i(x)$. The ansatz is more than a list of linear harmonics: it may contain nonlinear functions of $\phi^i$, gauge-covariant derivatives, warp factors, compensating fields, and fixed internal tensors.

Write the full parent equations, gauge constraints, and Bianchi identities collectively as $\mathcal E_A^{(D)}[\Phi]=0$. The minimal consistency statement is

$$\mathcal E_i^{(d)}[\phi]=0 \quad\Longrightarrow\quad \mathcal E_A^{(D)}[\mathcal U[\phi]]=0$$

for every allowed lower-dimensional solution. “Allowed” matters: the fields must obey the boundary conditions, global identifications, bundle data, flux sector, and regularity assumptions used to define the ansatz.

A particularly useful proof exhibits the stronger off-shell factorization

$$\mathcal E_A^{(D)}[\mathcal U[\phi]](x,y) =\mathcal K_A{}^i[\phi;y]\, \mathcal E_i^{(d)}[\phi](x),$$

where $\mathcal K_A{}^i$ may be a differential operator. All $y$-dependence then factors into known tensors, so the parent residual vanishes whenever the lower equations do. This is the structure realized by generalized Scherk–Schwarz reductions when their twist equations hold Hohm and Samtleben 2015, §§ 3.1–3.2.

For a scalar harmonic basis, a direct diagnostic is to project the residual onto every discarded mode $Y_I(y)$:

$$\mathcal S_I[\phi](x) =\int_K\!\mathrm d\mu_K\, Y_I^*(y)\, \mathcal E^{(D)}[\mathcal U[\phi]](x,y).$$

Consistency requires $\mathcal S_I[\phi]=0$ on every solution of the retained equations. Tensor fields require the appropriate inner product, gauge constraints, and representation projectors, but the logic is unchanged. A single nonzero $\mathcal S_I$ falsifies the proposed truncation.

A reproducible uplift test

The following workflow separates an exact embedding from an uncontrolled deletion.

1. Fix the parent problem. State the parent equations and derivative order, compact geometry, flux and bundle sector, boundary conditions, sources, and gauge or duality constraints. A proof for two-derivative supergravity is not automatically a proof after string or loop corrections.
2. Specify every retained field. Give its normalization, representation, and lower-dimensional gauge transformation. Include fields required by closure even if a preferred solution later sets them to zero.
3. Write the full nonlinear ansatz. Linearized mode functions are insufficient once products of retained fields generate new internal-coordinate dependence.
4. Reduce the retained equations. Insert the ansatz into the parent equations and identify the lower-dimensional equations, including constraints and topological terms.
5. Project every omitted equation. Test discarded harmonics and representations rather than only varying within the retained ansatz. Any nonzero source is a stop signal.
6. Cross-check the action and one uplift. When a lower action exists, verify that its Euler–Lagrange equations agree with the reduced equations. Then uplift a nontrivial lower solution and check the parent equations independently.
7. State the guarantee and its ceiling. Record whether the result is exact classical uplift, a perturbative or Wilsonian approximation, or only linear closure. Name the first omitted correction and a failure condition.

The expensive step is usually step 5. In a nonlinear sphere reduction, hundreds of apparently different internal tensors may have to reorganize into a small set of lower-dimensional equations. Symmetry can make the test short; exceptional geometry can make otherwise hidden factorization visible; neither can be replaced by a mass-gap slogan.

Exact worked example: the invariant circle mode

The simplest complete calculation already shows why symmetry and mass are different arguments. Work in Euclidean signature on $M_d\times S_R^1$, so $D=d+1$, with $\theta\in[0,2\pi)$ and

$$\begin{aligned} S_D=\int_{M_d}\!\mathrm d^dx\sqrt{g_d} \int_0^{2\pi}\!R\,\mathrm d\theta\, \bigg[ &\frac12(\nabla_d\Phi)^2 +\frac{1}{2R^2}(\partial_\theta\Phi)^2\\ &+\frac12M^2\Phi^2 +\frac{g}{3!}\Phi^3 \bigg]. \end{aligned}$$

The cubic potential is used only as a local closure diagnostic. Its equation of motion is

$$\mathcal E[\Phi] =\left(-\nabla_d^2-R^{-2}\partial_\theta^2+M^2\right)\Phi +\frac g2\Phi^2=0.$$

First retain only the circle-invariant mode,

$$\Phi(x,\theta)=\phi_0(x).$$

Every term in $\mathcal E[\phi_0]$ is independent of $\theta$. Orthogonality therefore makes every sine and cosine projection with nonzero harmonic number vanish identically. The only remaining equation is

$$\left(-\nabla_d^2+M^2\right)\phi_0 +\frac g2\phi_0^2=0.$$

It also follows from the unrescaled lower action

$$S_d=2\pi R\int_{M_d}\!\mathrm d^dx\sqrt{g_d}\, \left[ \frac12(\nabla_d\phi_0)^2 +\frac12M^2\phi_0^2 +\frac{g}{3!}\phi_0^3 \right].$$

This zero-mode truncation is exact for the stated classical scalar theory even if $R$ is not small. The reason is representation closure: the product of circle-translation singlets is again a singlet. It is not a claim about gravity backreaction or about a quantum completion of the toy potential.

Adversarial test: one retained harmonic sources another

Now try to retain a noninvariant cosine while discarding the rest:

$$\Phi(x,\theta)=\phi_0(x)+\phi_1(x)\cos\theta, \qquad \chi_2(x)=0.$$

Here $\chi_2$ denotes the coefficient that would multiply $\cos2\theta$ in the full harmonic expansion. The proposed truncation sets that independent field to zero.

The linear operator preserves harmonic number, but the interaction does not preserve this two-mode set because

$$\cos^2\theta=\frac12\left(1+\cos2\theta\right).$$

The projections onto the two retained harmonics give

$$\begin{aligned} 0={}&\left(-\nabla_d^2+M^2\right)\phi_0 +\frac g2\phi_0^2+\frac g4\phi_1^2,\\ 0={}&\left(-\nabla_d^2+M^2+R^{-2}\right)\phi_1 +g\phi_0\phi_1. \end{aligned}$$

These equations do not force $\phi_1$ to vanish.

Projecting the full equation onto $\cos2\theta$ gives

$$\frac1\pi\int_0^{2\pi}\!\mathrm d\theta\, \cos2\theta\,\mathcal E[\Phi] =\frac g4\phi_1^2.$$

Thus the discarded equation demands $g\phi_1^2/4=0$. It is not implied by the retained equations, so the finite set $\{1,\cos\theta\}$ is not a consistent truncation. This is the requested cubic-coupling falsifier: two factors of the retained $n=1$ cosine mode source the omitted $n=2$ harmonic.

If the $n=2$ field is restored through $\Phi\supset\chi_2(x)\cos2\theta$, its projected equation begins

$$\left(-\nabla_d^2+m_2^2+g\phi_0\right)\chi_2 +\frac g4\phi_1^2+\cdots=0, \qquad m_2^2=M^2+\frac4{R^2}.$$

The ellipsis includes couplings to other harmonics, all of which must also be matched if they are sourced. To first order in $g$ about $\phi_0=0$, and at momenta $Q^2\ll m_2^2$, solving rather than deleting $\chi_2$ gives

$$\chi_2 =-\frac g4\frac1{-\nabla_d^2+m_2^2}\phi_1^2 =-\frac{g}{4m_2^2} \left[1+O\!\left(\frac{Q^2}{m_2^2}\right)\right]\phi_1^2.$$

Substitution generates additional local interactions among the retained modes, followed by higher-derivative terms. The strongest surviving claim is therefore Wilsonian: after matching, low-energy observables can be accurate with a derivative-expansion remainder. Exact uplift with $\chi_2=0$ has failed. If $Q/m_2$ is not small, even that local approximation has no controlled truncation error.

Why a reduced action is not the proof

Substituting an ansatz into an action and then varying checks only variations tangent to the ansatz. Schematically,

$$\frac{\delta S_{\mathrm{red}}}{\delta\phi^i(x)} =\int_K\!\mathrm d\mu_K\, \left(\frac{\delta\mathcal U^A}{\delta\phi^i}\right)^{\!\dagger} \mathcal E_A^{(D)}[\mathcal U[\phi]].$$

This equation can vanish while an orthogonal discarded projection $\mathcal S_I$ remains nonzero. A plausible lower-dimensional action is therefore a valuable cross-check, but it is not by itself an uplift proof. Explicit Pauli sphere reductions show that even a consistent equation-level ansatz need not give the correct reduced theory when naively inserted into the parent action Lü, Pope, and Stelle 2007, abstract and §§ 3.1–3.2. The classic massless Kaluza–Klein gauge-field ansatz supplies the complementary failure: its discarded higher-dimensional equations are generically sourced Duff et al. 1984, pp. 90–94.

The reverse subtlety also occurs. Self-dual forms, dualized fields, boundary terms, and topological couplings may be simplest at the equation-of-motion level. One must show that any proposed lower action reproduces the already reduced equations with the correct duality prescription. In type-IIB supergravity, the self-duality equation $F_5=*F_5$ must be reduced together with the other parent equations; the $S^5$ construction checks it directly Baguet, Hohm, and Samtleben 2015, §§ 5.5 and 6. Thus five-form self-duality in this example is essential rather than optional.

Type IIB on the five-sphere

The important holographic application is type-IIB supergravity on $S^5$. For the complete bosonic sector at two-derivative order, the full nonlinear result is a consistent reduction to the bosonic sector of five-dimensional maximal $SO(6)$ gauged supergravity: every allowed bosonic solution of the five-dimensional theory has a ten-dimensional type-IIB uplift. Exceptional-field-theory twist equations make the factorization proof systematic, and explicit formulas reconstruct the ten-dimensional metric, scalars, form potentials—including the four-form potential—and the self-dual five-form field strength Baguet, Hohm, and Samtleben 2015, §§ 1–2 and 5–6. The authors explain that fermionic consistency follows from the supersymmetric framework, while their explicit original-IIB uplift formulas are bosonic.

This statement is stronger than the earlier explicit metric-plus-self-dual-five-form subsector. In that subsector, retaining the twenty $SL(6,\mathbb R)/SO(6)$ scalars while deleting all fifteen $SO(6)$ gauge fields is generically inconsistent: scalar currents source the gauge equations. Five-form self-duality also removes a would-be extra constraint Cvetič et al. 2000, § 2, Eqs. (6)–(10). These are concrete higher-dimensional versions of the circle diagnostic.

The uplift is nonlinear. A retained five-dimensional scalar can generate complicated $S^5$ dependence in the ten-dimensional metric and form fields. Calling the construction a truncation means that the coefficients of those internal tensors are fixed functions of the retained fields; it does not mean that every nonconstant harmonic literally disappears.

Within classical two-derivative type-IIB supergravity, the result licenses ten-dimensional uplifts of five-dimensional vacua, domain walls, black holes, and time-dependent solutions that obey the ansatz’s global and boundary conditions. It does not imply any of the following:

• every ten-dimensional solution lies in the retained sector;
• discarded fluctuations about an uplift are stable;
• the full Kaluza–Klein spectrum is represented;
• the path-integral measure or quantum theory truncates exactly; or
• $\alpha'$ effects, string loops, wrapped branes, or finite-$N$ effects preserve the same ansatz.

The last item is especially important in AdS$_5\times S^5$: the onset of the Kaluza–Klein tower has $m_{\mathrm{KK}}L=O(1)$. Consistency, not scale separation, licenses the classical five-dimensional sector. The D3 control analysis separately decides when that parent supergravity is a controlled approximation to string theory.

Four questions that must stay separate

Distinct tests for a finite lower-dimensional description

Question	Required test	What passing establishes	What it does not establish
Linear mode selection	The linearized operator preserves the retained harmonic subspace.	Independent propagation at first order around the chosen background.	Nonlinear closure or a finite interacting theory.
Consistent truncation	Every parent residual vanishes whenever the retained equations hold.	Exact classical uplift within the stated parent equations and global domain.	Completeness, stability, quantum exactness, or accuracy of the parent theory.
Wilsonian effective theory	Heavy fields are matched or solved for and the expansion parameters remain small.	Approximate low-energy observables with a declared remainder.	An exact finite-field uplift with the heavy fields set to zero.
Stability and completeness	Analyze all allowed fluctuations, global sectors, and omitted states.	Stability in the tested sector or coverage of a specified spectrum.	It is not supplied automatically by either consistency or EFT power counting.

Common pitfalls

“Keep the massless fields” is not a consistency proof. Interactions among massless modes can source massive harmonics. Test the discarded equations at nonlinear order.

A lower-dimensional action proves the uplift. Varying the reduced action probes only tangent variations. Orthogonal discarded equations can still fail.

A consistent truncation is a controlled approximation to the ultraviolet theory. Consistency is exact relative to specified parent equations. The reliability of those equations against string, loop, or higher-derivative corrections is a separate expansion problem.

Every higher harmonic is absent from a truncation. A nonlinear uplift may contain higher harmonics with coefficients fixed by retained fields. What is absent are independent lower-dimensional degrees of freedom for those shapes.

Exercises

1. Prove closure of every invariant polynomial interaction

Replace the cubic potential in the circle example by an arbitrary differentiable local potential $V(\Phi)$. Show that the ansatz $\Phi(x,\theta)=\phi_0(x)$ remains a consistent classical truncation.

Solution

The equation is

$$\left(-\nabla_d^2-R^{-2}\partial_\theta^2\right)\Phi +V'(\Phi)=0.$$

For $\Phi=\phi_0(x)$, both the kinetic term and $V'(\phi_0)$ are independent of $\theta$. Their projection onto every nonzero Fourier harmonic therefore vanishes by orthogonality. The remaining zero-mode equation is $-\nabla_d^2\phi_0+V'(\phi_0)=0$, exactly the equation obtained by substituting the ansatz into the action. Closure follows from circle-translation invariance, not from the numerical size of $R$.

2. Diagnose the reduced-action trap

Consider two $d$-dimensional scalar fields with

$$S=\int\!\mathrm d^dx\, \left[ \frac12(\partial\phi)^2 +\frac12(\partial\chi)^2 +\frac12m_\chi^2\chi^2 +g\chi\phi^2 \right].$$

Substitute $\chi=0$ into the action. Why does the resulting action not prove that retaining only $\phi$ is consistent? What low-energy claim survives for momenta well below $m_\chi$?

Solution

Substitution gives a free action for $\phi$, but the discarded equation is

$$\left(-\partial^2+m_\chi^2\right)\chi+g\phi^2=0.$$

At $\chi=0$ it requires $g\phi^2=0$, which is not implied by the free $\phi$ equation. The truncation is therefore inconsistent. For $Q^2\ll m_\chi^2$, one may instead solve

$$\chi=-\frac{g}{-\partial^2+m_\chi^2}\phi^2 =-\frac{g}{m_\chi^2} \left[1+O\!\left(\frac{Q^2}{m_\chi^2}\right)\right]\phi^2$$

and substitute this solution back. That produces a Wilsonian interaction and a derivative-expansion remainder; it does not make $\chi=0$ an exact uplift.

3. Audit an S⁵ uplift claim

A five-dimensional black-hole solution belongs to maximal $SO(6)$ gauged supergravity and obeys the conditions of the type-IIB uplift ansatz. Classify the following statements: (a) it has a ten-dimensional solution of two-derivative type-IIB supergravity; (b) it includes every Kaluza–Klein excitation; (c) it is stable against every ten-dimensional fluctuation; (d) it is an exact finite-$N$ string background.

Solution

Only (a) follows from consistency. Statement (b) confuses an embedded subsector with the full spectrum. Statement (c) requires a separate fluctuation analysis including discarded modes. Statement (d) additionally requires control of $\alpha'$ corrections, string loops, nonperturbative sectors, flux and global data, and the finite-$N$ dictionary. A consistent classical uplift supplies none of those conclusions automatically.

What the method licenses

The answer is now operational: a finite mode set is a consistent truncation only if the complete nonlinear ansatz makes every parent residual vanish on every retained solution in the declared global domain. A nonzero discarded-harmonic source ends that claim immediately. A large mass can still license an approximate lower-dimensional EFT, but only after the heavy field is solved for or matched and the remainder is controlled.

Volume XIV’s supergravity EFT treatment owns generic derivative and loop control. Volume X owns supersymmetric multiplet organization. Top-down, bottom-up, and UV-completion claims carries the exact-versus-effective distinction into model claims. Stringy and quantum corrections tests whether the two-derivative parent survives the next correction layer, while holographic matter and model building owns bottom-up deployment. None of those handoffs changes the uplift criterion established here.

The chapter overview contains the structure diagram and validity and failure diagram. They are embedded there once so that their shared chapter-level context is not repeated on every article.

For the chapter-wide comparison of assumptions, counterevidence, falsifiers, and claim ceilings, see the claim-domain table.

References

• Baguet, Arnaud, Olaf Hohm, and Henning Samtleben. “Consistent Type IIB Reductions to Maximal 5D Supergravity.” Physical Review D 92 (2015): 065004. DOI; arXiv.
• Cvetič, M., H. Lü, C. N. Pope, A. Sadrzadeh, and T. A. Tran. “Consistent $SO(6)$ Reduction of Type IIB Supergravity on $S^5$.” Nuclear Physics B 586 (2000): 275–286. DOI; arXiv.
• Duff, M. J., B. E. W. Nilsson, C. N. Pope, and N. P. Warner. “On the Consistency of the Kaluza–Klein Ansatz.” Physics Letters B 149 (1984): 90–94. DOI.
• Hohm, Olaf, and Henning Samtleben. “Consistent Kaluza–Klein Truncations via Exceptional Field Theory.” Journal of High Energy Physics 2015, no. 1 (2015): 131. DOI; arXiv.
• Lü, H., C. N. Pope, and K. S. Stelle. “Consistent Pauli Sphere Reductions and the Action.” Nuclear Physics B 782 (2007): 79–93. DOI; arXiv.