After a brief presentation of the exact renormalization group equation, we illustrate how the field theoretical (perturbative) approach to critical phenomena takes place in the more general Wilson (nonperturbative) approach. Notions such as the continuum limit and the renormalizability and the presence of singularities in the perturbative series are discussed.This paper has two parts. In the first part we restrict ourselves to the presentation of some selected issues taken from a recent review on the exact renormalization group (RG) equation (ERGE) in the pure scalar case.[1] In the second part we illustrate the Wilson continuum limit of field theory (or, equivalently, what may be understood as the nonperturbative renormalizability). More generally we would like to indicate how the historical first version of the RG, based on a perturbative approach, takes place in the more general nonperturbative framework developed by Wilson. The illustration will be done in the light of actual RG trajectories obtained from a numerical study of the ERGE in the local potential approximation. [2] 1 Selected issues
What does "exact" mean ?The word "exact" means: a continuous (i.e. not discrete) realization of the Wilson RG transformation of the action in which no approximation is made and no expansion is involved with respect to some small parameter. We are here only interested in the differential form of the ERGE. One could think that the word exact is not adapted to the Wilson renormalization because, compared to the standard version of renormalization 1 , it involves a finite cutoff and it is only a semi-group. Nevertheless the historical first version is entirely contained in the Wilson theory (see part 2).