This week I have been studying recent developments in Group Field Theories. Group field theories are a generalization of matrix models which provide both a second quantized reformulation of loop quantum gravity as well as generating functions for spin foam models. Other posts looking at this include:

- GFTcondensates and cosmology
- Quantum cosmology of loop quantum gravity condensates
- Constructing spacetime from the quantum tetrahedron: spacetime as a bose-einstein condensate
- Disappearance and emergence of spacetime in quantum gravity
- Numerical work using python 1: Bose-Einstein condensation

While states in canonical loop quantum gravity are based on graphs with vertices of arbitrary valence, group field theories have been defined so far in a simplicial setting such that states have support only on graphs of fixed valency. This has led to the question whether group field theory can indeed cover the whole state space of loop quantum gravity.

The paper discusses the combinatorial structure of the complexes generated by the group field theory partition function. These new group field theories strengthen the links between the various quantum gravity approaches and might also prove useful in the investigation of renormalizability.

**The combinatorial structure of group field theory**

The common notion of GFT is that of a quantum field theory on group manifolds with a particular kind of non-local interaction vertices. A group field is a function of a Lie group G and the GFT is defined by a partition function

the action is of the form:

The evaluation of expectation values of quantum observables O[φ], leads to a series of Gaussian integrals evaluated

through Wick contraction which are catalogued by Feynman diagrams Γ,

where sym(Γ) are the combinatorial factors related to the automorphism group of the Feynman diagram Γ:

The specific non-locality of each vertex is captured by a boundary graph. In the interaction term in each group field term can be represented by a graph consisting of a k-valent vertex connected to k univalent vertices. One may further understand the graph as

the boundary of a two-dimensional complex a with a single internal vertex v. Such a one-vertex two-complex a is called a spin foam atom.

The GFT Feynman diagrams in the perturbative sum have the structure of two complexes because Wick contractions effect bondings of such atoms along patches. The combinatorial

structure of a term in the perturbative sum is then a collection of spin foam atoms, one for each vertex kernel, quotiented by a set of bonding maps, one for each Wick contraction. Because of this construction such a two-complex will be called a spin foam molecule.

The crucial idea to create arbitrary boundary graphs in a more efficient way is to distinguish between virtual and real edges and obtain arbitrary graphs from regular ones by contraction of the

virtual edges.

In terms of these contractions, any spin foam molecule can be obtained from a molecule constructed from labelled regular graphs.

**Conclusions**

This paper has aimed to generalization of GFT to be compatible with LQG. It has clarified the combinatorial structure underlying the amplitudes of perturbative GFT using the notion of spin foam atoms and molecules and discussed their possible spacetime interpretation.

**Related articles**

- Comments of the Week #64: from quantum gravity to the weak force [Starts With A Bang] (scienceblogs.com)
- Foamy Evidence (briankoberlein.com)
- Feynman’s Fabulous Formula (golem.ph.utexas.edu)
- GFTcondensates and cosmology
- Quantum cosmology of loop quantum gravity condensates
- Constructing spacetime from the quantum tetrahedron: spacetime as a bose-einstein condensate
- Disappearance and emergence of spacetime in quantum gravity
- Numerical work using python 1: Bose-Einstein condensation