Matrix models and the gravitational interaction

The superstring theory is now regarded as the most promising candidate for the unification of the Standard Model and gravity, and this field has been rigorously investigated. However, we have seen a setback of the perturbative analysis of the superstring theory, because it has so many vacua that we have no way to determine which are the true ones. In order to remedy this problem and thus for the superstring theory to have a power to predict our real four-dimensional world, we need the nonperturbative formulation of the superstring theory. In the late 1990’s, our understanding of the nonperturbative aspects of the superstring theory have been greatly deepened. Especially, the large-N (N is the size of the matrices) reduced models have been proposed as the nonperturbative formulation of the superstring theory. One of the most promising candidates is the IIB matrix model, which is defined by the dimensional reduction of the ten-dimensional N = 1 super-Yang-Mills theory to zero dimension. It has been conjectured that this model resurrects the behavior of the string theory at the large-N limit. There have been a lot of interesting discoveries of the IIB matrix model, such as the dynamical generation of the four-dimensional spacetime and the interpretation of the diffeomorphism invariance. On the other hand, there are a lot of problems to surmount, if a large-N reduced model is to be an eligible framework to unify the gravitational interaction. Firstly, it is still an enigma how we can realize the local Lorentz invariant matrix model. In addition, we need to understand how we can describe the curved spacetime more manifestly, in terms of a large-N reduced model. This thesis discusses several attempts to address these issues concerning the gravitational interaction. This thesis is based on the following works [26,38,46,60].