Trans-Planckian Initial-State Sensitivity

“Trans-Planckian” sensitivity is not a universal prediction. Within effective field theory it is a dependence on boundary operators or modified dynamics suppressed by a physical cutoff, accompanied by running, stress-energy bounds, and truncation errors. Evolution above the cutoff has no model-independent output.

Required background. Initial-state boundary EFT supplies localized operators; vacuum choice fixes the reference state; and gravitational EFT power counting fixes the cutoff. Helpful background. Review Bunch–Davies and alpha diagnostics and controlled EFT expansion.

A boundary operator and its hierarchy

On an initial hypersurface $\Sigma_0$, consider the irrelevant operator

$$\delta S_0 =\frac{c_1(\mu)}{2\Lambda} \int_{\Sigma_0}d^3x\sqrt h\, h^{ij}D_i\phi D_j\phi.$$

It shifts the Gaussian boundary condition for a mode by

$$\delta\kappa_k\sim c_1(\mu)\frac{k_{\rm phys}^2(\eta_0)}{\Lambda}, \qquad k_{\rm phys}=\frac{k}{a_0}.$$

Perturbation theory requires $k_{\rm phys}/\Lambda\ll1$ for every mode used. Propagation converts this shift into a Bogoliubov correction and an oscillatory power-spectrum contribution. In common inflationary setups its envelope is of order $H/\Lambda$ times a Wilson coefficient, but the phase and even the leading power depend on the initial-surface prescription and operator basis.

Boundary loops renormalize $c_1$ and the lower-dimension boundary terms. Collins and Holman show why this running is required for initial-state predictions Collins and Holman 2005, §§II–IV.

First application: signal versus energy bound

Solve the mode equation to first order in $\delta\kappa_k$, compute

$$\frac{\delta\mathcal P_\phi(k)}{\mathcal P_\phi(k)} =2\operatorname{Re}\!\left(\frac{\delta v_k}{v_k}\right),$$

and retain only modes with $k/a_0\le\epsilon\Lambda$, $\epsilon\ll1$. Independently compute the renormalized excitation energy,

$$\delta\rho(\eta_0) \simeq\frac1{2\pi^2a_0^4} \int^{a_0\epsilon\Lambda}dk\,k^2\Omega_k|\beta_k|^2,$$

after the local vacuum subtraction. Require $\delta\rho\ll3M_{\rm Pl}^2H^2$ and that omitted boundary operators change the signal by less than the quoted uncertainty. Schalm, Shiu, and van der Schaar formulate this boundary-EFT organization for inflationary initial conditions Schalm, Shiu, and van der Schaar 2004, §§2–4.

What a controlled signal contains

A controlled result is a relation among low-energy observables and renormalized Wilson coefficients, not a determination of those coefficients from semiclassical evolution. Its prediction must include the retained operator basis, the symmetry assumptions that may suppress coefficients, the cutoff ratio over the fitted momentum window, and a bound on the first omitted order. A visually distinctive oscillation does not weaken these requirements.

The phase is particularly sensitive to preparation. A fixed initial time, a “new-physics hypersurface” imposed separately for each $k$, and a state prepared by earlier dynamics are inequivalent prescriptions. They can yield different functional dependence even at the same nominal power of $H/\Lambda$. Comparing amplitudes while suppressing this distinction converts a model choice into a false universal signature.

Field redefinitions and boundary equations of motion can also move contributions among operators. Observable predictions must remain invariant after the Wilson coefficients are translated. Finally, the excitation-energy test is necessary but not sufficient: a small integrated $\delta\rho$ can coexist with a narrow momentum band at the cutoff where the derivative expansion fails. Control must hold mode by mode over the reported domain as well as after integration.

The structure map places a trans-Planckian signature under both boundary power counting and backreaction checks.

Initial-state EFT can parameterize UV sensitivity through Wilson coefficients; it does not predict those coefficients or dynamics above the cutoff. Schematic; not to scale.

Domain and failure conditions

The chapter’s canonical domain table distinguishes a boundary-EFT correction from an alpha vacuum or UV completion. State the cutoff, initial slice, physical momentum window, operator basis, running scale, and stress bound.

Adversarial test. Move $\Sigma_0$ while representing the same physical state. Boundary coefficients and phases must run so the low-energy correlator is unchanged to the retained order. A leading feature that moves without compensation is preparation-surface dependence, not a trans-Planckian prediction. Likewise, a signal requiring $k_{\rm phys}\sim\Lambda$ lies outside the derivative expansion.

The failure map sends such a feature back to an explicit UV model rather than attaching a universal observational claim.

Slice independence after boundary running, cutoff suppression, and a small excitation stress are necessary conditions for a trans-Planckian EFT signature. Schematic; not to scale.

Evidence boundary

As of 10 August 2026, no model-independent trans-Planckian correction follows from semiclassical FLRW evolution. Durable claims are conditional Wilson-coefficient bounds and consistency relations; numerical amplitudes require a specified initial-state model and dated observational analysis.

References

• Collins, H., and R. Holman, “Renormalization of Initial Conditions and the Trans-Planckian Problem of Inflation,” Physical Review D 71, 085009 (2005), doi:10.1103/PhysRevD.71.085009.
• Schalm, K., G. Shiu, and J. P. van der Schaar, “Decoupling in an Expanding Universe: Boundary RG-Flow Affects Initial Conditions for Inflation,” Journal of High Energy Physics 2004, 076 (2004), doi:10.1088/1126-6708/2004/04/076.