so far all the responses seem to be more about a semantical distinction of the word evolution and natural selection
When we stress mice and notice a difference in their offspring through epigenetics, this is something that has already evolved in mice. Parts of the parental DNA get "marked" and the mark passes on to the offspring and influences their development, but the mechanism that connects external stress and the marking of the parental DNA is itself a genetic mechanism and subject to evolution.
The offspring whose development was changed because of the parental stress isn't evolved, much like a baby with FASD[1] isn't "evolved" either
[1]https://en.wikipedia.org/wiki/Fetal_alcohol_spectrum_disorde...