I'm reading chapter 4 of Peskin & Schröder, and I'm confused how they express the time ordering of two fields: let $T$ denote time ordering, $N$ the normal ordering and I use $C$ for the contraction of the two fields (as I couldn't find a way to write it better in mathjax)
$$T\{\phi(x)\phi(y)\}=N\{ \phi(x)\phi(y)+ C({\phi(x)\phi(y)})\} $$
My problem is that I think it should be:
$$T\{\phi(x)\phi(y)\}=N\{ \phi(x)\phi(y)\}+ C({\phi(x)\phi(y)}) $$
As $C({\phi(x)\phi(y)})$ is a commutator, and normal ordering a commutator will just result in $0$, more specifically, if for instance $y_0<x_0$
$$ C({\phi(x)\phi(y)})=\phi^+(x)\phi^-(y)-\phi^-(y)\phi^+(x)$$
where $\phi^\pm$ is the positive/negative frequency part of the field.
What am I missing? Moreover, they state that this $\phi^\pm$ decomposition is always possible for free fields, but I thought that the whole point of this was to deal with interacting fields, is it possible in general?