Loop quantum gravity: an outside view

TL;DR

This paper provides a critical, outsider's assessment of loop quantum gravity, arguing that a robust test of quantum spacetime covariance requires off-shell closure of the quantum constraint algebra. It emphasizes the many ambiguities in formulating the Hamiltonian constraint and the lack of a proven off-shell constraint algebra, which complicates establishing a reliable semiclassical limit and connection to standard particle physics. The authors contrast LQG’s finite, background-independent framework with perturbative gravity and string theory, highlighting both advances (spin networks, holonomies, area/volume spectra) and fundamental gaps (semi-classical states, regulator dependence, and full covariance). They advocate focusing on off-shell closure to constrain quantisation choices and reduce ambiguities, and call for further work to bridge LQG with low-energy physics and observable predictions.

Abstract

We review aspects of loop quantum gravity in a pedagogical manner, with the aim of enabling a precise but critical assessment of its achievements so far. We emphasise that the off-shell (`strong') closure of the constraint algebra is a crucial test of quantum space-time covariance, and thereby of the consistency, of the theory. Special attention is paid to the appearance of a large number of ambiguities, in particular in the formulation of the Hamiltonian constraint. Developing suitable approximation methods to establish a connection with classical gravity on the one hand, and with the physics of elementary particles on the other, remains a major challenge.

Figures (11)

• Figure 1: Setup used for the computation of the bracket (\ref{['e:basic_bracket']}). In the limit $\epsilon\rightarrow 0$ the edge $e(\epsilon)$ shrinks to zero and the two nodes just above and below the surface coincide.
• Figure 2: Examples of spin network states. For the 3-valent vertices on the left, the three incoming edges at each vertex are connected by a Clebsch-Gordan coefficient. For the 4-valent vertex on the right, one has to decide on a given way to construct a higher-order invariant tensor from two Clebsch-Gordan coefficients.
• Figure 3: A four-valent vertex is defined by a particular way of connecting three-valent vertices. The spin $k$ has to satisfy the triangle inequalities, but is otherwise arbitrary.
• Figure 4: A flux operator intersecting a three-valent node. The action is given by equation (\ref{['e:flux_3val_node']}).
• Figure 5: The action of the area operator on a node with intertwiner $C^{j_1 j_2 k}_{\alpha_1 \alpha_2 \beta} C^{j_3 j_4 k}_{\alpha_3 \alpha_4 \beta}$. In the figure on the left, the location of the edges with respect to the surface is such that the invariance of the Clebsch-Gordan coefficients (\ref{['e:CG_invariance']}) can be used to evaluate the action of the area operator. The result can be written in terms of a "virtual" edge. In the figure on the right, however, the edges are located in a different way with respect to the surface. The invariance property (\ref{['e:CG_invariance']}) does not apply, one has to use the recoupling relation (\ref{['e:recoupling']}), and the spin network state is therefore not an eigenstate of the corresponding area operator.