---
title: Semiclassical Stress Tensor Deformation
url: https://www.emergentmind.com/topics/semiclassical-stress-tensor-deformation-dictionary
type: topic
---

# Semiclassical Stress Tensor Deformation

The “semiclassical stress tensor deformation dictionary” (*Editor’s term*) denotes a family of constructions in which a deformation driven by the stress-energy tensor, or by operators built from it, is re-expressed as a change of background geometry, a mixed boundary condition, a quasi-local source–response relation, or a constitutive law. In the examples developed across field theory, holography, semiclassical gravity, condensed matter, and Madelung theory, the stress tensor is not merely an observable: it becomes the variable that organizes how geometry flows, how boundary data are related, or how effective sources are renormalized and interpreted [2410.02537], [2206.03415], [1603.04137], [2301.12170], [2606.09170], [1303.3582].

## 1. General structure of stress-tensor-driven deformation

A common starting point is a background-dependent action \(S_\tau[\phi,g]\) or \(A_\lambda[\Phi;g]\) together with the Hilbert or Brown–York stress tensor defined by metric variation. In the metric-flow formulation, the deformation is written as
\[
S_{\tau+\delta\tau}[\phi,g]
=
S_\tau[\phi,g]
+
\delta\tau\int d^d x\,\sqrt g\,\mathcal O_\tau(T_\tau),
\]
with
\[
T^{\mu\nu}_\tau
=
\frac{-2}{\sqrt g}\frac{\delta S_\tau}{\delta g_{\mu\nu}}.
\]
For polynomial stress-tensor deformations, the matter equations of motion are dynamically equivalent to evolving the background metric itself. One form of the dictionary is
\[
\frac{d g_{\mu\nu}}{d\tau}
=
2\,\frac{\mathcal O_\tau(T_\tau)}{T^{\mu\nu}_\tau},
\]
together with the coupled stress-tensor flow
\[
\frac{d T^{\mu\nu}_\tau}{d\tau}
=
\left(g^{\mu\nu}T^{\rho\sigma}_\tau-g^{\rho\sigma}T^{\mu\nu}_\tau\right)
\frac{\partial \mathcal O_\tau}{\partial T^{\rho\sigma}_\tau}
-
g^{\mu\nu}\mathcal O_\tau
-
2\frac{\partial \mathcal O_\tau}{\partial g_{\mu\nu}},
\]
and the invariant density
\[
\frac{d}{d\tau}\big(\sqrt g\,\mathcal O_\tau(T_\tau)\big)=0.
\]
The same idea appears in the higher-dimensional \(T\bar T\)-like bilinear family
\[
O_\lambda^{[r,d]}
=
\frac1d\left(r\,\mathrm{tr}[T_\lambda]^2-\mathrm{tr}[T_\lambda^2]\right),
\]
for which the deformation is dynamically equivalent to a field-dependent metric flow
\[
\frac{d g_{\mu\nu}}{d\lambda}
=
\pm\frac4d\,\big(r\,g_{\mu\nu}\,\mathrm{tr}[T_\lambda]-T_{\lambda,\mu\nu}\big).
\]
These formulations are explicitly classical or semiclassical and are stated as dynamical equivalences, not as general off-shell identities of functionals [2410.02537], [2206.03415].

Within this framework, exact solvable sectors play a special role. In \(d=2\), the \(T\bar T\) flow admits closed metric solutions, while for
\[
\mathcal O_\tau(T_\tau)=a_m(\mathrm{tr}\,T_\tau)^m
\]
the induced metric flow is a pure Weyl rescaling,
\[
g_{\mu\nu}(\tau)=\Omega(\tau)^2\,g_{\mu\nu}(0),
\qquad
\Omega(\tau)=
\left(
1+(m-1)d\,a_m(\mathrm{tr}[T_0])^{m-1}\tau
\right)^{\frac{m}{d(m-1)}}.
\]
This suggests that a large part of the dictionary can be read as a conversion between composite stress-tensor operators and effective geometry, although the papers are careful to formulate this at the classical level and, beyond special cases, to allow perturbative rather than exact solutions [2410.02537].

## 2. Finite-cutoff AdS and gravitational completions

A second line of development derives the dictionary directly from gravitational constraints. For Euclidean AdS gravity in Gaussian normal gauge,
\[
ds^2=dr^2+\gamma_{ij}(x,r)\,dx^i dx^j,
\qquad
K_{ij}=\frac12\partial_r\gamma_{ij},
\]
the Gauss–Codazzi Hamiltonian constraint implies a relation between the quasi-local stress tensor on a constant-radius slice and a specific quadratic stress-tensor composite. On flat constant-\(r\) slices, the central identity is
\[
-\,4\pi G\left(
T_{ij}T^{ij}-\frac{1}{d-1}(T^i{}_i)^2
\right)
=
T^i{}_i.
\]
This motivates the higher-dimensional deformation operator
\[
X_d
=
T_{ij}T^{ij}-\frac{1}{d-1}(T^i{}_i)^2,
\]
which reduces to the familiar \(2d\) structure when \(d=2\). When bulk gauge or scalar fields are present, the same radial-constraint analysis modifies the deformation by adding current-current and scalar-dependent terms, so the finite-cutoff dictionary is not purely quadratic in \(T_{ij}\) once extra bulk matter is turned on [1805.10287].

A more explicit gravitational formulation starts from
\[
S_{\mathrm{grav}}^{(\lambda)}[g,\psi]
=
\hat S[g,\psi]
+
\frac{l^2}{2\lambda}\int d^dx\,\sqrt g\,R,
\]
with metric saddle \(g^*_{\mu\nu}\) satisfying
\[
R_{\mu\nu}^*
-
\frac12 R^* g^*_{\mu\nu}
=
-\lambda l^{-2}\hat T^*_{\mu\nu}.
\]
The deformed QFT action associated with a saddle is then defined as the gravitational action evaluated on that saddle,
\[
S_\alpha^{(\lambda)}[\hat\gamma,\psi]
\equiv
S_{\mathrm{grav}}^{(\lambda)}[g_\alpha^*,\psi],
\]
with \(\hat\gamma_{\mu\nu}=\lim_{\lambda\to0}g^*_{\mu\nu}\). Around a fixed reference background \(\hat\gamma_{\mu\nu}\), the leading deformation is universal and bilocal, quadratic in the stress tensor with kernel set by the graviton Green’s function; on the saddle-point metric \(g^*_{\mu\nu}\), however, the flow becomes local because the metric has already absorbed the nonlocal gravitational response [2603.08481].

This split between a nonlocal fixed-background description and a local saddle-metric description is one of the sharpest formulations of the semiclassical dictionary in \(d>2\). A plausible implication is that the apparent nonlocality of higher-dimensional stress-tensor deformations is partly a choice of variables: locality is restored when the deformation is expressed on the metric selected by the gravitational equations of motion [2603.08481].

## 3. Boundary dynamics, mixed boundary conditions, and cosmological holography

In AdS/CFT with a dynamical boundary metric, the dictionary becomes a boundary semiclassical Einstein equation. The effective action is
\[
S_{\rm eff}
=
\frac{1}{16\pi G_d}\int d^d x\,\sqrt{-{\cal G}}\,({\cal R}-2\Lambda_d)
+
S_{\rm bulk}^{\rm on\mbox{-}shell}[{\cal G}],
\]
and variation with respect to the boundary metric \({\cal G}_{\mu\nu}\) yields
\[
{\cal R}_{\mu\nu}
-
\frac12 {\cal R}\,{\cal G}_{\mu\nu}
+
\Lambda_d{\cal G}_{\mu\nu}
=
8\pi G_d\,\langle T_{\mu\nu}\rangle.
\]
Here the holographic stress tensor is the renormalized Brown–York tensor,
\[
\langle T^{\mu\nu}\rangle
=
\frac{2}{\sqrt{-{\cal G}}}
\frac{\delta S_{\rm bulk}^{\rm on\mbox{-}shell}[{\cal G}]}
{\delta {\cal G}_{\mu\nu}},
\]
so the source of the boundary Einstein equation is the stress-tensor expectation value of the dual CFT itself. In asymptotic-mode language, the boundary gravitational self-action replaces Dirichlet holography by a mixed relation between slow and fast graviton falloffs. The parameter
\[
\gamma_d
=
\frac{G_d\,L}{\pi\,G_{d+1}}
\left(\frac{L}{\ell}\right)^{d-2}
\]
controls the size of the CFT contribution relative to the boundary cosmological term and fixes the boundary dynamics [2301.12170].

In dS/CFT, the corresponding proposal is formulated at future infinity \(\mathcal I^+\). The boundary deformation
\[
\frac{\partial S_\lambda[\gamma,\phi]}{\partial \lambda}
=
\int d^d x\,\sqrt{\gamma}\,\mathcal O
\]
induces coupled flow equations for the deformed boundary metric \(\gamma_{ab}\) and stress tensor \(T^{ab}\),
\[
\frac{\partial \gamma_{ab}}{\partial\lambda}
=
2\frac{\partial\mathcal O}{\partial T^{ab}},
\]
\[
\frac{\partial T^{ab}}{\partial\lambda}
=
\left(T^{cd}\gamma^{ab}-T^{ab}\gamma^{cd}\right)\frac{\partial\mathcal O}{\partial T^{cd}}
-
2\frac{\partial\mathcal O}{\partial\gamma_{ab}}
-
\mathcal O\,\gamma^{ab}.
\]
The holographic claim is that these field-theoretic flow equations are imposed as mixed boundary conditions for the bulk metric in asymptotically dS\(_3\) Fefferman–Graham form,
\[
ds^2
=
-\frac{\ell^2}{4\rho^2}d\rho^2
+
\frac1\rho\big(g^{(0)}_{ab}+\rho g^{(2)}_{ab}+\rho^2 g^{(4)}_{ab}\big)\,dx^a dx^b,
\]
with
\[
\gamma_{ab}=g^{(0)}_{ab},
\qquad
T_{ab}^{\rm BY}
=
\frac{1}{8\pi G\ell}
\left(
g^{(2)}_{ab}-g^{(2)c}{}_c\,g^{(0)}_{ab}
\right).
\]
In Kerr-dS\(_3\)/CFT\(_2\), the conserved charges constructed from the Brown–York tensor reproduce the boundary spectral flow exactly, and the same deformed geometry is used to compute pseudo entropy from complex geodesic saddles [2606.09170].

These constructions share a common structure. The stress tensor is simultaneously a response variable and a term in the dynamical equation for the metric, so the standard source/vev separation of holography is replaced by a mixed source–response relation. This suggests that “deformation by \(T_{\mu\nu}\)” is often more precisely a deformation of the boundary variational problem itself [2301.12170], [2606.09170].

## 4. Flat-space and spatial-infinity quasi-local dictionaries

A different holographic realization appears in the construction of a quasi-local stress tensor for asymptotically flat Kerr from Kerr–AdS. Starting from the Kerr–AdS metric in Boyer–Lindquist coordinates and the renormalized Brown–York tensor,
\[
T^{ij}
=
\frac{2}{\sqrt{-\gamma}}
\frac{\delta S}{\delta \gamma_{ij}},
\]
one finds that the naive flat limit \(\ell\to\infty\) of the AdS stress tensor does not exist componentwise: some components vanish and some diverge. The proposed cure is a Kerr-specific deformation dictionary motivated by Flat/CCFT. After first shifting to a non-rotating boundary frame,
\[
\phi=\varphi-\frac{a}{\ell^2}t,
\]
one rescales stress-tensor components by appropriate powers of \(\ell/\sqrt G\) before taking the limit. The resulting asymptotically flat stress tensor has nonzero components
\[
8\pi\,\tau_{tt}=\frac{2M}{r\sqrt G},
\qquad
8\pi\,\tau_{t\varphi}
=
-\frac{3aM}{r\sqrt G}\sin^2\theta,
\]
\[
8\pi\,\tau_{\theta\theta}
=
\frac{M\sqrt G}{r},
\qquad
8\pi\,\tau_{\varphi\varphi}
=
\frac{M\sqrt G}{r}\sin^2\theta.
\]
The dual boundary geometry is obtained not at null infinity but at spatial infinity, after the ultra-relativistic contraction
\[
\tilde t=\frac{\sqrt G}{\ell}\,t,
\]
which turns the boundary metric into a finite contracted geometry [1603.04137].

The Brown–York charge formula
\[
Q_\xi
=
\int_\Sigma d\sigma\,\sqrt{\sigma}\,v^\mu \xi^\nu \tau_{\mu\nu}
\]
then yields
\[
Q_{\partial_t}=M,
\qquad
Q_{\partial_\varphi}=-aM,
\]
which are the Kerr mass and angular momentum up to the sign convention for the rotation generator. The same stress tensor is also checked from the CCFT side using the BMS\(_4\)-type generators
\[
L_n,\quad \bar L_n,\quad M_{m,n},
\]
and the corresponding charge relations. The paper is explicit that this is not yet a generally covariant flat-space renormalization theorem; the construction depends on Boyer–Lindquist coordinates, on a non-rotating boundary frame, and on the contracted geometry at spatial infinity rather than the usual null-infinity setting [1603.04137].

The significance of this example is methodological. It shows that in four dimensions a stress-tensor deformation dictionary may require anisotropic contraction of boundary coordinates, componentwise rescaling of Brown–York data, and charge matching as the test of correctness, rather than a raw limit of an already-renormalized AdS tensor [1603.04137].

## 5. Constitutive and material-geometric dictionaries

In several non-holographic settings, the stress tensor is also tied to deformation through a constitutive dictionary. In Madelung theory, the usual stress tensor
\[
\sigma_{ij}
=
\frac{\hbar^2}{4m}\rho\,\partial_i\partial_j(\ln \rho)
=
\frac{\hbar^2}{2m}\left(
R\,\partial_i\partial_jR-\partial_iR\,\partial_jR
\right)
\]
is not interpreted as the stress of a deformed material body. Instead, the deformation is a deformation of the local coframe,
\[
\theta^i=R\,dx^i,
\qquad
g_{ij}=\rho\,\delta_{ij},
\]
and the strain variable is the one-form
\[
\omega
=
\frac1R\,dR
=
d(\ln R)
=
\frac12 d(\ln \rho).
\]
The constitutive law becomes
\[
\sigma
=
\frac{\hbar^2}{2m}\rho\,d\omega.
\]
This is explicitly a gradient constitutive law for frame strain, not a Hooke-type algebraic law and not an ordinary fluid-viscous law. The same \(\omega\) is also interpreted as the teleparallel connection of the deformed frame, with
\[
d\theta^i=-\omega\wedge\theta^i,
\qquad
\nabla\theta^i=0,
\qquad
\Omega=d\omega+\omega\wedge\omega=0.
\]
The paper therefore replaces ordinary material strain by “frame strain,” and rewrites Madelung stress as the derivative of that frame strain [1303.3582].

In deforming crystals, the geometry is organized by a lattice bundle over spacetime. The local lattice vectors \(a_\alpha^i(\boldsymbol x,t)\) determine the reciprocal basis \(b^\alpha_i\) and the lattice connection
\[
\Gamma^i_{j\mu}
=
b^\alpha_j\,\partial_\mu a_\alpha^i.
\]
This connection encodes both strain gradient and strain rate. The appropriate derivative is the lattice-covariant derivative
\[
\nabla_{x^\mu}f
=
\left(
\partial_{x^\mu}
-
k_l\Gamma^l_{j\mu}\partial_{k_j}
\right)f,
\qquad
Dk_i=dk_i+k_j\Gamma^j_{i\mu}dx^\mu,
\]
and the deformation potential is the covariant strain derivative of the band energy,
\[
\nabla_{x^\mu}\varepsilon
=
D^n_m\,\Gamma^m_{n\mu}.
\]
The semiclassical equations become
\[
D_t\boldsymbol x
=
\partial_{\boldsymbol k}\varepsilon_{\mathrm{tot}}
-
D_t\boldsymbol k\times\boldsymbol\Omega_{\boldsymbol k}
-
\boldsymbol\Omega_{\boldsymbol kT},
\]
\[
D_t\boldsymbol k
=
-
\nabla_{\boldsymbol x}\varepsilon_{\mathrm{tot}}
+
m_e D_t\boldsymbol x\times 2\boldsymbol\omega
-
m_e\boldsymbol a,
\]
and the electron energy stress tensor for a spatially homogeneous band insulator is
\[
\mathcal T^i{}_j
=
\mathcal D^i{}_j
+
2\omega\cdot\nabla_j^{\,i}\mathcal J
+
\Gamma_{m0}^{\,n}\eta_{nj}^{\,mi}
+
a\cdot\nabla_j^{\,i}P_{m_e}
+\cdots.
\]
Here the stress is explicitly conjugate to strain, strain rate, rotation, and acceleration through deformation-potential, viscosity, angular-momentum, and mass-polarization structures [1802.02887].

These material examples broaden the scope of the dictionary. They show that a “stress-tensor deformation dictionary” need not always mean deforming an action by \(T\bar T\)-like composites; it may instead mean identifying the precise geometric variable whose derivative or curvature the stress tensor measures [1303.3582], [1802.02887].

## 6. Solvable model classes and explicit deformed observables

The most explicit solvable examples occur in \(2d\) \(T\bar T\)-type theories and in \(4d\) nonlinear electrodynamics. In the random-geometry approach to \(2d\) \(T\bar T\), the infinitesimal deformation is written as a Gaussian average over random metric perturbations \(h_{ij}\),
\[
e^{-\delta S}
\propto
\int [dh]\,
\exp\left[
-\frac{1}{8\delta\mu}\int d^2x\,\sqrt g\,\epsilon^{ik}\epsilon^{jl}h_{ij}h_{kl}
-\frac{1}{4\pi}\int d^2x\,\sqrt g\,h_{ij}T^{ij}
\right].
\]
At the saddle,
\[
h_{ij}
=
-\delta\mu\,\epsilon_{ik}\epsilon_{jl}T^{kl},
\]
so the deformed theory on a fixed background is reinterpreted as an undeformed theory averaged over nearby geometries. The first-order correction to the Polyakov action is
\[
\delta S[g]
=
-\frac{c^2\delta\mu}{48\pi}
\int d^2x\,\sqrt g\,
R\,
\frac1{\Box+\frac R2}
\nabla_k
\frac1{\Box+\frac R2}
\nabla^kR.
\]
This generates the deformed stress-tensor correlators. A salient structural result is that the first-order deformation leaves two-point functions unchanged, modifies three-point and four-point functions, and produces logarithmic corrections beginning at the four-point level, not in the two- or three-point sector [2012.03972].

In \(4d\) nonlinear electrodynamics, the distinguished \(T\bar T\)-analogue is
\[
O_{T^2}
=
\frac18\left(
T^{\mu\nu}T_{\mu\nu}
-
\frac12 (T^\mu{}_\mu)^2
\right).
\]
The paper proves that, among irrelevant stress-tensor deformations of \(4d\) Abelian electrodynamics, this is the unique one preserving zero birefringence. For a Lagrangian \(\mathcal L(S,P)\), the flow is
\[
\frac{\partial \mathcal L}{\partial\lambda}
=
-\frac12\Big[
(\mathcal L-P\mathcal L_P)^2
-
2S(\mathcal L-P\mathcal L_P)\mathcal L_S
-
P^2(\mathcal L_S)^2
\Big].
\]
The associated root-\(T^2\) operator is
\[
R
=
\frac12\sqrt{
T^{\mu\nu}T_{\mu\nu}
-
\frac14(T^\mu{}_\mu)^2
}
=
\sqrt{S^2+P^2}\,\mathcal L_S,
\]
and its flow generates ModMax-like families. Within this dictionary, Born–Infeld is the \(T^2\) trajectory from Maxwell, Plebanski is a classical \(T^2\)-fixed point, and reverse Born–Infeld is an analytically continued or subtracted trajectory. The same paper also constructs manifestly \(\mathcal N=1\) supersymmetric superspace versions of both the \(T^2\)-like and root-\(T^2\)-like flows [2301.10411].

These solvable models show two recurrent patterns. First, the deformation is often simplest when written as a flow of couplings or metric data rather than as an operator insertion problem. Second, special theories may become fixed points or acquire constant stress-tensor-squared operators after subtraction, indicating that the stress-tensor dictionary can reorganize a theory into a more geometric variable set [2012.03972], [2301.10411].

## 7. Renormalization, exceptional source classes, and limits of semiclassicality

Any semiclassical stress-tensor dictionary is constrained by renormalization and by the validity of the semiclassical approximation itself. In curved-spacetime QFT, the renormalized one-point function is defined by point splitting and Hadamard subtraction,
\[
\langle \psi | T_{ab}(x) | \psi \rangle_{\rm ren}
=
\lim_{x\to x'}
D_{ab'}
\Big[
\langle \psi | \phi(x)\phi(x') | \psi \rangle
-
H(x,x')
\Big],
\]
but \(\langle T_{ab}\rangle_{\rm ren}\) is only defined up to conserved local geometric counterterms. For quartic composites such as \(T_{ab}T_{cd}\), state-independent Hadamard subtraction is no longer sufficient; the renormalized local second moment must be defined through OPE-renormalized four-point functions. The paper then formulates a necessary semiclassicality criterion in tetrad components,
\[
\left|
\langle T_{IJ}^2(x)\rangle_{\rm ren}
\right|
\le
\epsilon\,\langle T_{IJ}(x)\rangle_{\rm ren}^2,
\qquad
\epsilon<1.
\]
Squeezed vacua, including those relevant to black-hole evaporation and inflationary cosmology, fail this criterion: the renormalized fluctuations are generically of the same order as the mean. This places a direct restriction on any dictionary that uses only \(\langle T_{ab}\rangle_{\rm ren}\) as source data [2512.17789].

A related issue is algebraic classification. Type III stress tensors,
\[
T_{ab}^{(\mathrm{III}_0)}
=
f(\ell_a s_b+s_a\ell_b),
\qquad
\ell^2=0,\quad s^2=1,\quad \ell\cdot s=0,
\]
satisfy
\[
(T^2)^a{}_b=f^2\ell^a\ell_b,
\qquad
(T^3)^a{}_b=0.
\]
They are therefore nilpotent of index \(3\), non-diagonalizable, and structurally fragile. The paper emphasizes that only types I and IV are stable under generic perturbations, whereas types II and III are not. This matters for deformation theory because a stress-tensor dictionary built on algebraic type must distinguish robust sectors from measure-zero Jordan degenerations [1907.01269].

Semiclassical backreaction can also be constructive rather than only restrictive. In static spherically symmetric stellar equilibrium with a constant-density classical fluid plus vacuum polarization in the Boulware state, the semiclassical Einstein equation
\[
G_{\mu\nu}
=
8\pi\left(
T_{\mu\nu}
+
\langle \hat T_{\mu\nu}\rangle
\right)
\]
turns the classical isotropic star into an effectively anisotropic source. With the regularized Polyakov approximation, the renormalized stress-energy tensor introduces deformation of the effective density, radial pressure, and tangential pressure, and a minimal deformation of the Polyakov approximation inside a tiny central core is sufficient to produce regular ultracompact stars beyond the Buchdahl limit. In this setting, the stress-tensor deformation dictionary takes the form of an effective-source replacement
\[
\rho_{\rm eff}=\rho+\rho_{\rm q},
\qquad
p_{r,\rm eff}=p+p_{r,\rm q},
\qquad
p_{t,\rm eff}=p+p_{t,\rm q},
\]
which modifies the stellar equilibrium equations without changing the classical conservation law of the fluid itself [2110.15808].

Taken together, these results delimit the domain of the subject. A semiclassical stress-tensor deformation dictionary can be exact, constructive, and geometrically illuminating, but it is never independent of renormalization scheme, fluctuation control, coordinate or boundary choice, and the algebraic structure of the stress tensor itself [2512.17789], [1907.01269], [2110.15808].

Source: https://www.emergentmind.com/topics/semiclassical-stress-tensor-deformation-dictionary