Causal Dynamical Triangulations in Quantum Gravity

• Causal Dynamical Triangulations (CDT) is a nonperturbative method that discretizes spacetime using foliated, causal triangulations built from Minkowskian simplices.
• It exhibits a rich phase structure with distinct geometric phases and phase transitions that recover semiclassical de Sitter space and support a continuum limit.
• CDT integrates matter fields and quantum observables to explore Planck-scale phenomena within a background-independent framework of quantum gravity.

Causal Dynamical Triangulations (CDT) is a nonperturbative, lattice-based approach to quantum gravity designed to provide a rigorous definition of the gravitational path integral by summing over discretized, causal spacetime histories. CDT incorporates a global, discrete time foliation and strictly enforces causal structure by restricting summation to foliated triangulations built from flat Minkowskian simplices. The formalism aims to recover macroscopic spacetime physics compatible with general relativity from a background-independent, quantum sum over geometries, while probing novel Planck-scale quantum gravitational phenomena unobtainable in perturbative or Euclidean frameworks.

1. Discrete Formulation and Path Integral

The foundation of CDT is the replacement of the continuum gravitational path integral

$$Z(G, \Lambda) = \int \mathcal{D}[g] e^{i S_{\mathrm{EH}}[g]}$$

(where $S_{\mathrm{EH}}[g]$ is the Einstein–Hilbert action) by a regularized sum over piecewise-flat triangulations $T$ that admit a global, discrete time foliation. Each $d$-dimensional triangulation is constructed from two types of $(d+1)$-simplices distinguished by their vertex distribution across adjacent time slices: in four dimensions, (4,1) and (3,2) simplices. Spacetime is sliced into $T$ layers labeled by integer proper time, with each spatial slice $\Sigma(t)$ carrying a fixed topology—commonly $S^3$ or $T^3$.

Upon Wick rotation (analytically continuing the Lorentzian signature parameter $\alpha \to -\alpha$), the partition function becomes a sum over Euclidean triangulations:

$S_{\mathrm{EH}}[g]$0

where $S_{\mathrm{EH}}[g]$1 is the automorphism group order and $S_{\mathrm{EH}}[g]$2 is a Regge-type discretization of the Einstein–Hilbert action:

$S_{\mathrm{EH}}[g]$3

with $S_{\mathrm{EH}}[g]$4 the vertex count, $S_{\mathrm{EH}}[g]$5 and $S_{\mathrm{EH}}[g]$6 the counts of four-simplices of each type, $S_{\mathrm{EH}}[g]$7 the inverse Newton coupling, $S_{\mathrm{EH}}[g]$8 the cosmological coupling, and $S_{\mathrm{EH}}[g]$9 the asymmetry parameter encoding the ratio of space- to time-like edge lengths. All spatial topology change is forbidden in standard CDT, ensuring manifoldness and causal ordering at each step (Loll, 2019, Ambjørn et al., 7 Apr 2026, Ambjorn et al., 2013).

2. Phase Structure and Critical Phenomena

CDT exhibits a rich phase diagram in the $T$0 bare coupling space, characterized by distinct geometric phases and a hierarchy of phase transitions:

• A phase (“branched polymer”): Spatial slices are uncorrelated; geometry is dominated by disconnected, polymeric branches; no extended semiclassical geometry forms.
• B phase (“crumpled” or collapsed): Triangulation accumulates most volume into few slices or vertices, yielding infinite spectral and Hausdorff dimensions.
• C phase and its subphases (“de Sitter—C$T$1”, “bifurcation—C$T$2”): An extended semiclassical universe with a continuous volume profile emerges, consistent on large scales with a four-sphere (Euclidean de Sitter). The C$T$3 phase is marked by bifurcation of the transfer matrix’s kinetic term and singular vertices with high coordination (Gizbert-Studnicki, 2017, Ambjorn et al., 2021).

The order of these transitions is crucial for continuum physics. The A–C transition is first order, while the B–C and C$T$4–C$T$5 transitions are established as second order (or higher), supporting the existence of divergent correlation lengths and the possibility of a well-defined continuum limit. Critical exponents from finite-size scaling and Binder cumulant analysis confirm these orders; e.g., at the C$T$6–B transition on toroidal topology, $T$7 (Ambjorn et al., 2021), while typical second-order signatures at B–C include exponents $T$8 (Ambjorn et al., 2012, Ambjorn et al., 2019).

Phase structure and critical phenomena are robust against changes in spatial topology: moving from $T$9 to $d$0 slices leaves both the nature of the phases and the order of transitions unchanged, with only minor quantitative shifts in critical points, demonstrating universality (Gizbert-Studnicki, 2019, Ambjorn et al., 2019).

3. Emergent Geometry and Quantum Observables

In the physically relevant C$d$1 phase, CDT realizes a dynamically generated four-dimensional universe with semiclassical properties:

• Spatial Volume Profile: The ensemble average of the three-volume per slice, $d$2, exhibits the universal form $d$3 characteristic of a Euclidean de Sitter solution. This result is quantitatively reproduced by an effective one-dimensional minisuperspace action derived via transfer-matrix measurements:

$d$4

with $d$5 a quantum correction manifest specifically for toroidal topology (Ambjorn et al., 2021, Ambjorn et al., 2013, Trzesniewski, 2011).

• Spectral Dimension and Dimensional Reduction: The spectral dimension $d$6, defined via diffusion processes on spatial slices, interpolates between $d$7 at macroscopic scales and $d$8 in the ultraviolet, strongly indicating dynamical dimensional reduction at the Planck scale (Clemente et al., 2019, Mielczarek, 2015, Loll, 2019).
• Fractal Properties and Loop Lengths: The minimal non-contractible loop length scales as $d$9, indicating a semiclassical core with superposed fractal outgrowths; the static Hausdorff dimension is measured at $(d+1)$0 in the semiclassical phase (Ambjorn et al., 2021).
• Harmonic Scalar Probes: On toroidal topology, massless scalar fields with prescribed winding are used as geometric probes. Classical solutions to the Laplacian serve as pseudo-coordinates, revealing the fractal and void-and-filament structure of the quantum-generated geometry. For subcritical winding, the scalar is a passive probe; above a critical winding $(d+1)$1, the geometry exhibits pinching, leading to a matter-induced localization transition (Ambjorn et al., 2021).

4. Matter Coupling and Extended Observables

CDT admits the inclusion of additional fields as dynamical degrees of freedom. The partition function generalizes to

$(d+1)$2

where $(d+1)$3 is the discretized scalar action incorporating winding number. Scalar fields with topological winding can induce nontrivial geometric backreaction, driving new quantum phase transitions such as the formation of quantum "necks"—localized regions capturing nearly all nontrivial winding (Ambjorn et al., 2021).

Inclusion of matter, such as Yang–Mills fields, follows a gauge-invariant discretization using Wilson loops on the dual lattice. In 2D, minimal coupling of U(1) and SU(2) gauge fields leaves gravitational critical exponents unaffected, but the fluctuating geometry substantially decorrelates long-distance topological gauge modes (Candido et al., 2021).

5. Finite-Size Scaling, Continuum Limit, and Universality

Establishing a rigorous continuum limit is tied to the existence of higher-order (second or greater) transitions, where the correlation length diverges as $(d+1)$4. Key numerically measured exponents include:

Transition	Order	Critical Exponent $(d+1)$5	Reference
C$(d+1)$6–A	First	$(d+1)$7	(Ambjorn et al., 2012)
C$(d+1)$8–B	Second+	$(d+1)$9	(Ambjorn et al., 2021)
B–C	Second	$T$0	(Ambjorn et al., 2012)

Topology does not affect these universal critical properties; phase diagrams, critical exponents, and observable spectra are coincident (up to finite-size lattice artifacts) between $T$1 and $T$2, underlining universality (Gizbert-Studnicki, 2019).

Second-order lines in the coupling space are natural candidates for UV fixed points, where the CDT lattice formulation aims to define a continuum quantum field theory of gravity, potentially connected to asymptotic safety (Ambjorn et al., 2013, Loll, 2019).

6. Outlook and Open Challenges

CDT has established the robust emergence of semiclassical four-dimensional spacetime from Planckian quantum dynamics, provided evidence for ultraviolet critical points, and offered a computational playground for exploring minimal gravitational path integrals with causality. Key open challenges highlighted in recent research include:

• Identification and classification of local, diffeomorphism-invariant observables sensitive to Planckian geometry and their cosmological implications (Ambjørn et al., 7 Apr 2026).
• Precise characterization of the continuum limit, particularly at the triple-point intersection C$T$3–C$T$4–B, and its relation to universality classes such as Hořava–Lifshitz gravity (Ambjorn et al., 2021).
• Analytical understanding of novel matter-induced phase transitions and fractal structure revealed by harmonic coordinate probes on toroidal spatial topology.
• Derivation and interpretation of genuinely quantum terms in effective actions (e.g., the $T$5 term for toroidal geometry).
• Extension to more realistic matter couplings with nontrivial topology, and defining effective cosmological scenarios from nonperturbative Planck-scale quantum spacetime (Ambjorn et al., 2021).

CDT remains a leading nonperturbative quantum gravity framework, providing direct access to emergent spacetime phases, critical geometry transitions, and the interplay between topology, matter, and quantum microstructure (Ambjorn et al., 2021, Loll, 2019, Gizbert-Studnicki, 2019).