Timelike-Boundary Causality and Boundary-Value Problems

Evolution in AdS is well posed only after initial data are supplemented by boundary data that define an appropriate self-adjoint evolution—or by an explicit open-system completion. For a closed scalar system, admissible reflective or mixed conditions make the renormalized symplectic flux vanish. Locality in boundary time is also consequential: an arbitrary frequency-dependent relation can conserve neither flux nor causal support.

Required background. Anti-de Sitter geometry supplies the timelike conformal boundary. Timelike boundaries and self-adjoint extensions supplies the general operator-domain analysis. Helpful background. Covariant symplectic structure supplies the conserved Klein–Gordon pairing used below.

Why initial data are insufficient

Take a real scalar in Lorentzian global AdS$_{d+1}$ with action

$$S=\frac12\int_M\mathrm d^{d+1}x\sqrt{\lvert g\rvert} \left(g^{MN}\partial_M\phi\,\partial_N\phi-m^2\phi^2\right).$$

With signature $(+,-,\ldots,-)$ the equation is $(\Box+m^2)\phi=0$. A radial null ray reaches the timelike boundary in finite global time, so a spacelike slice is not a Cauchy surface for the spacetime without boundary data. In a static treatment, choosing a positive self-adjoint extension of the spatial wave operator specifies the allowed dynamics under the hypotheses stated by Ishibashi and Wald 2003, §§2–3.

Near $z=0$, write

$$\phi=z^{\Delta_-}\alpha+z^{\Delta_+}\beta+\cdots, \qquad \nu=\frac{\Delta_+-\Delta_-}{2}.$$

For two linearized solutions, the radial symplectic flux through a boundary region is proportional to

$$\mathcal F_{\partial M}(\phi_1,\phi_2) =2\nu\int_{\partial M}\!\mathrm d^d x\sqrt{\lvert g_{(0)}\rvert} \left(\alpha_1\beta_2-\beta_1\alpha_2\right),$$

after the required renormalization at exceptional masses. Dirichlet data $\delta\alpha=0$, alternate data $\delta\beta=0$ when allowed, or a real local mixed relation $\beta=W'(\alpha)$ make the antisymmetrized flux vanish. This condition is the covariant counterpart of symmetry of the spatial operator; positivity of the extension is an additional stability requirement Ishibashi and Wald 2004, §§3–4.

First application: a reflective global-AdS scalar problem

Fix standard quantization and $\alpha=0$ for normalizable fluctuations. Separation of variables gives

$$\phi_{n\ell m}=e^{-i\omega_{n\ell}\tau} f_{n\ell}(\rho)Y_{\ell m}(\Omega), \qquad \omega_{n\ell}=\Delta_++2n+\ell.$$

Regularity at $\rho=0$ selects one radial solution, while $\alpha=0$ selects the discrete frequencies. The Klein–Gordon pairing on a global slice $\Sigma_\tau$ is

$$(\phi_1,\phi_2)_{\mathrm{KG}} =i\int_{\Sigma_\tau}\mathrm d\Sigma^M \left(\phi_1^*\nabla_M\phi_2-\phi_2\nabla_M\phi_1^*\right).$$

Stokes’ theorem says that its change between two slices equals minus the boundary flux. Because $\alpha=0$, that flux vanishes and the pairing is conserved. Regular interior data plus a reflective boundary condition therefore yield a reproducible normal-mode evolution. This does not establish an interacting quantum theory or a dual CFT.

Causality and open boundary conditions

A boundary relation may be flux conserving yet problematic if it is nonlocal in time. A kernel

$$\beta(t,\mathbf x)=\int\mathrm dt'\,K(t-t')\alpha(t',\mathbf x)$$

requires a causal support condition on $K$ and a compatible enlarged Hilbert space. A generic real function $\beta(\omega)=f(\omega)\alpha(\omega)$ is not automatically the boundary value of a causal response: its analytic structure must obey the relevant retarded prescription, and any absorbed flux must be carried by declared boundary degrees of freedom.

“Transparent” boundary conditions therefore describe a closed unitary system only after the exterior or bath that receives the flux is included. Within the AdS region alone they define an open sector. Bulk causality is then assessed in the combined system, not by pretending the boundary was reflective.

Adversarial check: an uncontrolled frequency-dependent condition

Let $\beta(\omega)=i\gamma\omega\alpha(\omega)$ with real $\gamma>0$. For a real-time mode, the boundary flux contains an absorptive term proportional to $\gamma\omega^2\lvert\alpha\rvert^2$. The AdS-region Klein–Gordon norm is not conserved. Formal solutions of the radial equation still exist, but their frequencies need not be real and they do not define a unitary closed-system Hamiltonian.

If instead $f(\omega)$ is chosen arbitrarily with poles in the upper half-plane, the inverse kernel has support before the source and violates retarded causality. The strongest surviving claim is only that one has solved a differential equation with a specified nonlocal boundary relation. A well-posed causal quantum evolution requires self-adjointness or an explicit causal open-system completion, positive energy where claimed, and controlled boundary response.

Controlled limits and handoff

The reflective calculation is linear and fixed-background; it does not prove nonlinear stability or construct the exterior degrees of freedom required by transparent data. Asymptotically Locally AdS Boundary Data generalizes the source and response coefficients, while Boundary Conditions, Alternate Quantization, and Deformations classifies the BF-window mixed conditions in detail.

Exercise

Show directly that a real linear mixed condition $\beta=f\alpha$, with constant $f$, has zero antisymmetrized symplectic flux.

Solution

Substituting $\beta_i=f\alpha_i$ gives $\alpha_1\beta_2-\beta_1\alpha_2=f\alpha_1\alpha_2-f\alpha_1\alpha_2=0$. Reality of $f$ is needed for the Hermitian pairing. Vanishing flux does not by itself prove positivity; a negative enough mixed coupling can create an unstable bound state.

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

• Ishibashi, Akihiro, and Robert M. Wald. “Dynamics in Non-Globally-Hyperbolic Static Spacetimes II: General Analysis of Prescriptions for Dynamics.” Classical and Quantum Gravity 20 (2003): 3815–3826. arXiv. DOI.
• Ishibashi, Akihiro, and Robert M. Wald. “Dynamics in Non-Globally-Hyperbolic Static Spacetimes III: Anti-de Sitter Spacetime.” Classical and Quantum Gravity 21 (2004): 2981–3014. arXiv. DOI.