Strong de Sitter Condition

• Strong de Sitter Condition is a context‐dependent quantitative criterion that refines de Sitter statements across quantum gravity, scalar potentials, and geometric analyses.
• In swampland EFT and asymptotic compactification, it forbids metastable de Sitter vacua by imposing sharp gradient and curvature constraints, linking accelerated expansion to higher-dimensional de Sitter focus.
• It also serves as a diagnostic and constructive tool in black-hole physics and higher-derivative gravity, setting thresholds for cosmic censorship and de Sitter stability.

Across the cited literature, the expression “Strong de Sitter Condition” labels several non-equivalent but structurally related criteria. In swampland-oriented work it denotes a sharpened obstruction to positive scalar potentials or metastable de Sitter vacua; in asymptotic compactification analysis it denotes the claim that asymptotic accelerated expansion requires higher-dimensional metastable de Sitter and simultaneously loses Kaluza–Klein scale separation; in black-hole physics it denotes a quantitative threshold controlling strong cosmic censorship or timelike focusing in de Sitter-related geometries; and in constructive or higher-derivative settings it can instead denote a sufficient condition for de Sitter existence or a stability criterion for de Sitter fixed points (Obied et al., 2018, Andriot et al., 2018, Hebecker et al., 2023, Liu et al., 2019, Borissova et al., 10 Sep 2025, Rummel et al., 2011, Do, 25 May 2026). This suggests that the phrase is best understood as a context-dependent strengthening of a de Sitter-related statement rather than as a single universally standardized conjecture.

1. Terminological scope

The common feature of these usages is not a unique formula but a shared role: each “strong” condition sharpens a weaker de Sitter statement by imposing a quantitative criterion on a potential, a decay rate, a compactification asymptotic, or a curvature inequality.

Context	Strong condition	Representative source
Swampland EFT	$|\nabla V| \ge c\,V$ for $V>0$	(Obied et al., 2018)
Refined swampland EFT	$$\left(M_{\rm Pl}\frac{|\nabla V|}{V}\right)^q-a\,M_{\rm Pl}^2\frac{\min(\nabla_i\nabla_jV)}{V}\ge b$$	(Andriot et al., 2018)
Asymptotic compactification	asymptotic acceleration $\Rightarrow$ higher-dimensional metastable de Sitter and $M_{\rm KK}<H$	(Hebecker et al., 2023)
RN–dS / KdS SCC	$\beta=-{\rm Im}(\omega_0)/\kappa_-$ with threshold $1/2$	(Liu et al., 2019, Dias et al., 2018)
Regular black holes	$-m''(r)\ge 0$ as the TCC-specific inequality	(Borissova et al., 10 Sep 2025)
De Sitter no-hair	DEC + SEC + expansion + 2D-orbit symmetry force late-time de Sitter behavior	(Creminelli et al., 2020)
Higher-derivative or constructive settings	nonlinear ODE, Kähler-uplift window, or higher-curvature stability criterion	(Krishnan, 2019, Rummel et al., 2011, Do, 25 May 2026)

A recurrent misconception is to treat the phrase as synonymous with the original swampland gradient bound. The literature instead shows several domain-specific meanings. In some papers the strong condition forbids metastable de Sitter; in others it diagnoses SCC violation, timelike-convergence failure, or even provides a sufficient condition for constructing a metastable de Sitter vacuum.

2. Swampland formulations in scalar effective field theory

In the swampland literature, the strongest early formulation is the pointwise gradient inequality

$|\nabla V| \ge c\cdot V ,$

imposed wherever the scalar potential is positive, with $c$ a positive constant of order one in Planck units (Obied et al., 2018). For canonically normalized scalars with kinetic term $V>0$0, the gradient norm is

$V>0$1

Because a metastable de Sitter vacuum requires $V>0$2 and $V>0$3, the inequality directly forbids stable or metastable de Sitter vacua. The proposal is motivated by scaling arguments and no-go theorems in M-theory compactifications, the non-supersymmetric heterotic $V>0$4 string, Type II flux compactifications with orientifolds and D-branes, and bounds derived from the strong or null energy conditions (Obied et al., 2018).

That proposal also constrains accelerated cosmology. For a single canonical field with an exponential potential in four dimensions, the equation-of-state asymptotes to $V>0$5, so acceleration requires $V>0$6; correspondingly, if the universal bound $V>0$7 holds wherever $V>0$8, any accelerating solution requires $V>0$9 in four dimensions (Obied et al., 2018). In this usage, “strong” means universal, pointwise, and purely first-derivative: the criterion depends only on the slope and not on the Hessian.

A later refinement combines first and second derivatives into a single covariant inequality,

$$\left(M_{\rm Pl}\frac{|\nabla V|}{V}\right)^q-a\,M_{\rm Pl}^2\frac{\min(\nabla_i\nabla_jV)}{V}\ge b$$0

or equivalently

$$\left(M_{\rm Pl}\frac{|\nabla V|}{V}\right)^q-a\,M_{\rm Pl}^2\frac{\min(\nabla_i\nabla_jV)}{V}\ge b$$1

Here $$\left(M_{\rm Pl}\frac{|\nabla V|}{V}\right)^q-a\,M_{\rm Pl}^2\frac{\min(\nabla_i\nabla_jV)}{V}\ge b$$2 and $$\left(M_{\rm Pl}\frac{|\nabla V|}{V}\right)^q-a\,M_{\rm Pl}^2\frac{\min(\nabla_i\nabla_jV)}{V}\ge b$$3 (Andriot et al., 2018). At a positive critical point, this reduces to

$$\left(M_{\rm Pl}\frac{|\nabla V|}{V}\right)^q-a\,M_{\rm Pl}^2\frac{\min(\nabla_i\nabla_jV)}{V}\ge b$$4

so any de Sitter critical point must contain a tachyonic direction, excluding metastable de Sitter vacua. The paper derives this condition in the weak-coupling, semi-classical regime from the Swampland Distance Conjecture, entropic arguments involving Gibbons–Hawking entropy and the Bousso bound, and the control assumption $$\left(M_{\rm Pl}\frac{|\nabla V|}{V}\right)^q-a\,M_{\rm Pl}^2\frac{\min(\nabla_i\nabla_jV)}{V}\ge b$$5 (Andriot et al., 2018).

The same paper tests the inequality against classical Type IIA data and extracts the indicative parameter range

$$\left(M_{\rm Pl}\frac{|\nabla V|}{V}\right)^q-a\,M_{\rm Pl}^2\frac{\min(\nabla_i\nabla_jV)}{V}\ge b$$6

with $$\left(M_{\rm Pl}\frac{|\nabla V|}{V}\right)^q-a\,M_{\rm Pl}^2\frac{\min(\nabla_i\nabla_jV)}{V}\ge b$$7 (Andriot et al., 2018). Within that framework, concave slow-roll inflation is easily accommodated, whereas standard exponential quintessence is disfavored because the inequality permits only $$\left(M_{\rm Pl}\frac{|\nabla V|}{V}\right)^q-a\,M_{\rm Pl}^2\frac{\min(\nabla_i\nabla_jV)}{V}\ge b$$8, while the observational tension quoted in the paper is $$\left(M_{\rm Pl}\frac{|\nabla V|}{V}\right)^q-a\,M_{\rm Pl}^2\frac{\min(\nabla_i\nabla_jV)}{V}\ge b$$9 (Andriot et al., 2018).

A different quantum-gravity strengthening appears in the $\Rightarrow$0-matrix approach. There the relevant condition is not a slope bound but the requirement that any de Sitter-like phase exit before its quantum break-time,

$\Rightarrow$1

with

$\Rightarrow$2

For de Sitter, the corpuscular picture gives $\Rightarrow$3, shortened by light species and, in string theory, by finite $\Rightarrow$4 (Dvali, 2020). Exact de Sitter would require $\Rightarrow$5, which in this framework occurs only in the trivializing limit $\Rightarrow$6. This formulation is then mapped to swampland-type constraints on scalar potentials and to the bound on the duration of slow roll.

3. Asymptotic acceleration, decompactification, and higher-dimensional de Sitter

A distinct strong de Sitter condition is formulated in the study of asymptotic accelerated expansion in $\Rightarrow$7-dimensional EFTs. The central conjecture is Asymptotic Acceleration Implies de Sitter (AA$\Rightarrow$8DS): if a $\Rightarrow$9-dimensional theory exhibits accelerated expansion from rolling scalars at asymptotically large distance in field space, then it must arise from compactification of a $M_{\rm KK}<H$0-dimensional theory with positive vacuum energy, namely a higher-dimensional metastable de Sitter vacuum; moreover, in any such asymptotic accelerating regime the Kaluza–Klein scale eventually becomes lighter than the Hubble scale (Hebecker et al., 2023).

This statement sharpens the strong asymptotic de Sitter conjecture, or “No Asymptotic Acceleration,” whose $M_{\rm KK}<H$1-dimensional gradient bound is

$M_{\rm KK}<H$2

in $M_{\rm KK}<H$3-dimensional Planck units (Hebecker et al., 2023). The logic is explicitly dimensional: if the higher-dimensional de Sitter conjecture holds and no metastable de Sitter vacuum exists above dimension $M_{\rm KK}<H$4, then asymptotic acceleration in $M_{\rm KK}<H$5 dimensions is also excluded. In this sense, higher-dimensional de Sitter structure controls lower-dimensional asymptotics.

The argument proceeds by identifying asymptotic infinite-distance regions with decompactification limits. For $M_{\rm KK}<H$6, the Emergent String Conjecture motivates the claim that such limits bring down KK towers and can be viewed as decompactification of at least some internal directions (Hebecker et al., 2023). In a decompactification regime $M_{\rm KK}<H$7 with $M_{\rm KK}<H$8 internal directions of common radius $M_{\rm KK}<H$9, the positive terms in the $\beta=-{\rm Im}(\omega_0)/\kappa_-$0-dimensional potential scale as

$\beta=-{\rm Im}(\omega_0)/\kappa_-$1

These terms all dilute too rapidly to sustain asymptotic acceleration except for the higher-dimensional vacuum energy, equivalently a spacetime-filling $\beta=-{\rm Im}(\omega_0)/\kappa_-$2-brane or a $\beta=-{\rm Im}(\omega_0)/\kappa_-$3-form flux (Hebecker et al., 2023).

In Einstein frame, the canonically normalized volume modulus $\beta=-{\rm Im}(\omega_0)/\kappa_-$4 obeys

$\beta=-{\rm Im}(\omega_0)/\kappa_-$5

The flattest positive exponential then comes from the higher-dimensional cosmological constant,

$\beta=-{\rm Im}(\omega_0)/\kappa_-$6

which is strictly less than the acceleration threshold

$\beta=-{\rm Im}(\omega_0)/\kappa_-$7

By contrast, the next-flattest source, a codimension-1 brane with $\beta=-{\rm Im}(\omega_0)/\kappa_-$8, is always too steep for asymptotic acceleration (Hebecker et al., 2023).

The strong condition also includes an asymptotic loss of scale separation. The paper shows that

$\beta=-{\rm Im}(\omega_0)/\kappa_-$9

so maintaining $1/2$0 asymptotically requires

$1/2$1

Equivalently, whenever

$1/2$2

the KK scale drops below the Hubble scale asymptotically (Hebecker et al., 2023). Since this inequality is stricter than the condition for accelerated expansion,

$1/2$3

every asymptotic accelerating solution, and even some slightly steeper ones, fails $1/2$4-dimensional scale separation at late times. For $1/2$5, the flattest potential saturates the asymptotic TCC slope

$1/2$6

4. Black-hole formulations: strong cosmic censorship in de Sitter backgrounds

In de Sitter black-hole physics, the phrase is tied to Strong Cosmic Censorship rather than to swampland scalar potentials. The relevant quantity is

$1/2$7

for the dominant quasinormal mode, or more generally $1/2$8 in terms of the spectral gap $1/2$9 and the Cauchy-horizon surface gravity $-m''(r)\ge 0$0 (Liu et al., 2019, Dias et al., 2018). The Christodoulou criterion is the local Sobolev threshold $-m''(r)\ge 0$1 at the Cauchy horizon. For scalar perturbations, $-m''(r)\ge 0$2 implies local square-integrability of the gradient and hence possible weak extendibility across the Cauchy horizon; $-m''(r)\ge 0$3 blocks such weak extension.

For higher-dimensional Reissner–Nordström–de Sitter black holes, neutral massless scalar perturbations exhibit three dominant QNM families: photon-sphere modes, de Sitter modes, and near-extremal modes. The near-extremal family has

$-m''(r)\ge 0$4

for the dominant $-m''(r)\ge 0$5 mode, so $-m''(r)\ge 0$6 near extremality (Liu et al., 2019). The numerical conclusion is that in $-m''(r)\ge 0$7 there is always a near-extremal charge region with $-m''(r)\ge 0$8, so SCC is violated in the Christodoulou sense. The dependence on dimension is non-monotonic and organized into four regimes of $-m''(r)\ge 0$9: for example, Region I obeys $|\nabla V| \ge c\cdot V ,$0, while Region IV obeys $|\nabla V| \ge c\cdot V ,$1 in the sense of how difficult it is to violate SCC (Liu et al., 2019).

Kerr–de Sitter exhibits the opposite behavior. There, for every non-extremal black hole, there exist photon-sphere quasinormal modes with sufficiently slow decay that $|\nabla V| \ge c\cdot V ,$2, for both scalar and linearized gravitational perturbations (Dias et al., 2018). The eikonal branch is

$|\nabla V| \ge c\cdot V ,$3

so

$|\nabla V| \ge c\cdot V ,$4

throughout the subextremal parameter space, with saturation only in the extremal limit (Dias et al., 2018). Consequently, Christodoulou’s weak SCC holds for all non-extremal Kerr–de Sitter black holes, in sharp contrast to near-extremal RN–dS.

In this black-hole usage, the “strong de Sitter condition” is therefore a regularity threshold at the Cauchy horizon, quantified by the competition between exterior exponential decay and interior blueshift. It is conceptually unrelated to the swampland gradient bound, despite the shared phrase.

5. Geometric focusing and de Sitter no-hair

Another usage arises from timelike focusing in regular black holes with de Sitter or anti–de Sitter cores. For a spherically symmetric metric

$|\nabla V| \ge c\cdot V ,$5

the timelike convergence condition decomposes into three independent inequalities on the Misner–Sharp mass: $|\nabla V| \ge c\cdot V ,$6 In stationary spacetimes only the first two remain, and only

$|\nabla V| \ge c\cdot V ,$7

is genuinely TCC-specific; the second is already implied by the null convergence condition (Borissova et al., 10 Sep 2025). This inequality is the paper’s “Strong de Sitter Condition.”

The near-core expansion

$|\nabla V| \ge c\cdot V ,$8

distinguishes de Sitter from anti–de Sitter cores through the sign of $|\nabla V| \ge c\cdot V ,$9. For a smooth de Sitter core with approximately constant $c$0,

$c$1

so $c$2 is violated for sufficiently small $c$3 (Borissova et al., 10 Sep 2025). By contrast, an anti–de Sitter core with $c$4 satisfies the inequality locally near the core, but asymptotic flatness and positive ADM mass force a violation at some finite radius away from the core. The Bardeen black hole explicitly realizes global NCC satisfaction together with near-core TCC violation, while a modified Bardeen black hole with an anti–de Sitter core realizes local NCC violation near the core and TCC-specific violation in an annulus away from it (Borissova et al., 10 Sep 2025).

A different geometric strong condition appears in the de Sitter no-hair theorem for $c$5-dimensional cosmologies with positive cosmological constant and spatial slices foliated by closed two-dimensional symmetry orbits. Under DEC and SEC, a compact Cauchy surface, an everywhere expanding initial slice with $c$6, only crushing singularities, and either orbit Euler characteristic $c$7 or sufficiently large initial $c$8 orbits quantified by an explicit $c$9, the future region admits a global mean-curvature-flow foliation and becomes physically indistinguishable from de Sitter on arbitrarily large regions (Creminelli et al., 2020). The theorem gives exponential control of the mean curvature toward

$V>0$00

exponential convergence of the spatial metric to an exponentially expanding de Sitter-like comparison metric, exponential convergence of timelike and null lengths to their de Sitter values, and $V>0$01 dilution of matter in the causal domain of late-time observers (Creminelli et al., 2020).

These geometric usages differ from swampland statements in a crucial way. They do not primarily constrain scalar potentials. Instead, they formulate “strong de Sitter” as a focusing, censorship, or asymptotic-geometry condition inside black holes or in expanding cosmologies with $V>0$02.

6. Higher-derivative, duality-invariant, and constructive realizations

In $V>0$03-covariant, all-order $V>0$04 cosmology, the strong condition is a necessary differential constraint for Einstein-frame de Sitter. The Hohm–Zwiebach cosmological equations are controlled by a single even function $V>0$05 of the string-frame Hubble parameter, with

$V>0$06

and supergravity limit

$V>0$07

String-frame de Sitter requires the algebraic conditions

$V>0$08

at some nonzero $V>0$09. Einstein-frame de Sitter is stronger: it requires the second-order nonlinear ODE

$V>0$10

which sharply restricts admissible $V>0$11 (Krishnan, 2019). The paper further shows that constant-dilaton cosmologies are only power-law,

$V>0$12

so no constant-dilaton de Sitter exists in this framework, and simple analytic power-series expansions around $V>0$13 consistent with the supergravity limit cannot satisfy the strong ODE (Krishnan, 2019).

A very different usage is constructive rather than prohibitive. In Type IIB Kähler uplifting, a metastable de Sitter vacuum exists in the large-volume, weak-coupling regime if the dimensionless combination

$V>0$14

lies in the narrow window

$V>0$15

with sign and parametric conditions

$V>0$16

The extremum condition is

$V>0$17

Within this window the minimum is metastable, has $V>0$18, and persists under inclusion of the dilaton and complex-structure moduli; a structurally similar sufficient condition extends to swiss-cheese geometries (Rummel et al., 2011). This is a notable reversal of the swampland usage: the strong condition here is a sufficient criterion for realizing, rather than excluding, metastable de Sitter.

Higher-curvature gravity provides yet another variant. In generalized Einsteinian cubic gravity with an added $V>0$19 term,

$V>0$20

the exact de Sitter background obeys

$V>0$21

and only the cubic density $V>0$22 contributes to the value of $V>0$23 (Do, 25 May 2026). Without $V>0$24, the perturbed dynamical system at the de Sitter fixed point is incomplete because $V>0$25 is undetermined. Adding $V>0$26 closes the system; the Jacobian eigenvalues become

$V>0$27

For the Starobinsky choice $V>0$28, one eigenvalue is positive and the de Sitter fixed point is unstable, whereas the tuned second-order case

$V>0$29

has a stable de Sitter background fixed point with single eigenvalue $V>0$30 (Do, 25 May 2026). The paper explicitly notes that this background-level stability does not by itself yield a scalar-EFT counterexample to strong de Sitter swampland criteria, because a full perturbative ghost and gradient analysis is not performed.

Taken together, these higher-derivative and constructive examples show that the phrase can denote a severe local ODE constraint, a narrow sufficient window for metastable de Sitter vacua, or an instability criterion for exact de Sitter fixed points. The term therefore spans prohibition, diagnosis, and construction, depending on the theoretical setting.