about the action S in quantum field theory

[eljose79]

Let,s suppose we have a Hamiltonian H so we can construct the action by H+dS/dt then why no use the action to solve the problem of quantization of non renormalizable theories?..

[jeff]

I think you probably meant ∂S/∂t + H = 0. In any event, nonrenormalizability is no longer viewed as a problem. A theory is renormalizable if it's lagrangian need only contain a finite number of terms to absorb the different types of divergences that occur in it's perturbation theory. But now we know that it's perfectly alright to allow theories which require an infinite number of terms as long as their contribution to the theory are suppressed at higher energies.