Validity, Breakdown, and Quantum-Gravity Handoffs

The boundary of curved-spacetime QFT is not a single energy marked “Planckian.” Validity depends on what is being predicted, at what resolution and duration, in which state, with which curvature and coupling hierarchy, and to what tolerance. This final chapter supplies a disciplined way to stop, downgrade, or transfer a claim while preserving every low-energy result that remains controlled.

Helpful background. Fixed-background, semiclassical, gravitational-EFT, and quantum-gravity regimes supplies the regime distinctions; gravitational-EFT validity and breakdown supplies cutoff logic; stochastic-gravity validity limits supplies fluctuation diagnostics; and quasi-de Sitter validity supplies finite-duration and evidence boundaries.

A validity verdict is attached to an observable

For a specified observable $\mathcal O$, collect the calculation into an error contract

$$\mathcal C_{\mathcal O}= \left(\mathcal O,\rho,f,\mathcal D; \epsilon_{\rm curv},\epsilon_E,\epsilon_{\rm loop}, \epsilon_{\rm sec},\epsilon_{\rm fluc};\tau_{\mathcal O}\right).$$

Here $\rho$ is the state, $f$ represents smearing or detector resolution, $\mathcal D$ is the causal domain and duration, the $\epsilon_i$ are dimensionless expansion or fluctuation measures, and $\tau_{\mathcal O}$ is the required accuracy. The prediction is controlled only if omitted terms are bounded below $\tau_{\mathcal O}$ and the mathematical construction exists on $\mathcal D$.

Different observables on the same geometry can receive different verdicts. A renormalized mean stress may be reliable while its unsmeared pointwise variance is undefined; a detector response may remain finite while a global particle number is ambiguous; a low-frequency Hawking flux may be robust while the endpoint geometry is outside the derivative expansion. Stochastic gravity makes this distinction concrete by treating the noise kernel as a distribution whose operational consequences require smearing and a response problem Hu and Verdaguer 2020, §§3–5, Eqs. (3.1)–(5.28).

Likewise, gravitational EFT remains predictive below its cutoff even though it is not a microscopic completion. Nonanalytic long-distance quantum corrections can be universal while analytic local terms remain Wilson coefficients to be matched Donoghue 1994, §§II–IV, Eqs. (2.1)–(4.13).

The structure map separates three outcomes: continue within the approximation, downgrade to a narrower observable or shorter domain, or transfer a precisely posed remainder.

Validity belongs to an observable-specific error contract. A failed control parameter narrows or transfers only the affected claim while preserving independent low-energy results. Schematic; not to scale.

Routes through the chapter

1. Observable-specific validity contracts assigns separate scales and tolerances to each predicted quantity.
2. Semiclassical breakdown diagnostics combines mean equations, linear response, noise, state, and EFT tests.
3. Cross-expansion hierarchies determines whether curvature, derivatives, loops, occupation, or duration fails first.
4. Large-N, species, and cutoff hierarchies distinguishes fixed bare from fixed renormalized limits.
5. Trans-Planckian sensitivity compares robustness claims across horizons and cosmology.
6. Cauchy and chronology horizons identifies failures of global construction and state propagation.
7. Singularities and initial conditions separates observable breakdown from universal endpoint claims.
8. Evaporation endpoints and information marks controlled semiclassical intervals and unresolved completion questions.
9. Low-energy constraints on UV completion states qualified causality, analyticity, and positivity bounds.
10. Analogue and phenomenological evidence follows measurements only as far as their model map licenses.
11. The quantum-gravity handoff packages an unresolved observable without selecting a favored framework.

Domain and failure conditions

This is the canonical comparison table for the chapter. Mutable evidence claims are bounded by information available through 10 August 2026.

Problem	Observable and domain	Control data	Strongest licensed output	Failure or destination
Validity contract	Declared smeared or relational object	State, resolution, duration, tolerance	Object-specific error bound	Narrow the object or improve the calculation
Semiclassical background	Mean geometry plus selected probes	Equation residual, response spectrum, noise kernel	Stable mean-field prediction with fluctuation estimate	Stochastic treatment, higher EFT order, or transfer
Cross-expansion ordering	One finite-duration process	Curvature, energy, coupling, loop, occupation, secular parameters	First omitted-order estimate	Resum, include operators, or shorten the domain
Large-N and species	Renormalized gravitational theory with $N$ fields	Limit held fixed, thresholds, running couplings	Parametric mean and fluctuation hierarchy	Match thresholds or lower the gravitational cutoff
Trans-Planckian robustness	Hawking flux or inflationary correlator	State, dispersion, preferred frame, adiabaticity	Conditional low-energy insensitivity	Unknown UV initial or propagation data
Cauchy or chronology horizon	Algebra, state, and smeared observable in a stated region	Extension, wavefront set, stress response	Region-limited theorem or divergence statement	Non-globally-hyperbolic boundary data or rigorous treatment
Singular limit	Detector, stress, or correlator approaching a boundary	Curvature, state, smearing, extension	Prediction up to a stated stopping surface	Initial-condition or quantum-gravity remainder
Evaporation endpoint	Flux, geometry, entropy, or information proxy	Adiabaticity, curvature, fluctuations, entropy prescription	Controlled interval of evaporation	Endpoint and microscopic unitarity question
UV consistency bound	On-shell low-energy amplitude or invariant coefficient	Analyticity, causality, subtractions, massless exchanges	Qualified Wilson-coefficient constraint	UV model comparison after IR completion
Analogue or astrophysical evidence	Directly measured spectrum or propagation observable	Calibration, nuisance model, map fidelity	Kinematic or EFT parameter statement	Gravitational dynamics or microscopic completion remains open
Final transfer	Exact unresolved observable and surviving constraints	Failed approximation, error budget, alternatives, evidence grade	Framework-neutral problem statement	Quantum-gravity or rigorous destination

The validity map emphasizes that high curvature is only one path to failure. Long duration, large occupation, state singularity, global causal failure, uncontrolled fluctuations, or an inference step beyond measured data can intervene first.

There is no universal breakdown number: each failure channel is compared with the resolution and tolerance of one observable, and the surviving claims remain valid within their own domains. Schematic; not to scale.

References

• Donoghue, J. F., “General Relativity as an Effective Field Theory: The Leading Quantum Corrections,” Physical Review D 50, 3874–3888 (1994), doi:10.1103/PhysRevD.50.3874.
• Hu, B. L., and E. Verdaguer, “Stochastic Gravity: Theory and Applications,” Living Reviews in Relativity 23, 3 (2020), doi:10.1007/s41114-020-00026-7.