Spectral Dimension and Dimensional Flow

Spectral dimension measures how a fictitious diffusion process returns to its starting point. It is a scale-dependent probe of a chosen Laplacian and ensemble, not a direct measurement of causal, Hausdorff, or topological dimension. Similar ultraviolet values in different programs do not establish a shared theory.

Required background. Elliptic Boundary Problems and Heat Kernels supplies the continuum definition. Causal Dynamical Triangulations: Phases and Continuum Evidence supplies the main application.

Helpful background. Causal-Set Kinematics and Dynamics and Tensor Models and Random Geometry supply alternative ensembles.

Return probability and estimator

For positive Laplacian $\Delta$, the heat kernel obeys

$$\partial_\sigma K(x,x';\sigma)=-\Delta K(x,x';\sigma), \qquad K(x,x';0)=\delta(x,x').$$

The volume-averaged return probability and spectral dimension are

$$P(\sigma)=\frac1V\operatorname{Tr}e^{-\sigma\Delta}, \qquad D_s(\sigma)=-2\frac{d\log P}{d\log\sigma}.$$

On flat $\mathbb R^d$, $P=(4\pi\sigma)^{-d/2}$ and $D_s=d$.

Application: a discrete fit

On each triangulation, evolve a random walk with transition matrix $T$ and estimate

$$P(\sigma)=\frac1{N}\sum_{i=1}^{N}(T^\sigma)_{ii}.$$

Fit $\log P$ locally against $\log\sigma$ using correlated errors over the geometry and walk samples. Reject very small $\sigma$, where lattice parity and coordination dominate, and large $\sigma$, where finite volume drives $P\to1/N$ and $D_s\to0$. Vary the window and lattice volume.

CDT simulations found a long-distance value near four and a short-distance flow toward about two in the studied phase Ambjørn, Jurkiewicz, and Loll 2005. This is evidence for scale-dependent diffusion on that ensemble, not proof that every quantum-gravity ultraviolet limit is two-dimensional.

Cross-program comparison

In a causal set, a Lorentzian retarded d’Alembertian is not automatically a positive diffusion generator; any Euclideanized or absolute operator must be declared. In tensor graphs, the walk probes graph adjacency and may diagnose branched-polymer geometry. In an FRG calculation, an anomalous propagator can imply a return-probability scaling only under assumptions about the effective operator and measure.

Adversarial window and operator tests

Change the diffusion kernel, laziness parameter, ensemble weighting, fit window, and boundary treatment. Verify the result on a known $d$-dimensional lattice at matched volume. Compare $D_s$ with independently measured Hausdorff and causal dimensions. A flow that disappears under these controls is an estimator artifact; a robust flow still identifies only diffusion behavior.

Scope

Spectral dimension is useful because it is portable, but that portability comes from discarding causal and much geometric information. It is one row, not the conclusion, in Emergence and Continuum-Limit Tests.

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

• Ambjørn, J., J. Jurkiewicz, and R. Loll. “Spectral Dimension of the Universe.” Physical Review Letters 95 (2005): 171301. DOI.
• Carlip, S. “Dimension and Dimensional Reduction in Quantum Gravity.” Classical and Quantum Gravity 34 (2017): 193001. DOI.