verb classes

in my recent linguistics reading i've come across empirical methods for classifying verbs into clusters based on the contexts in which they occur. essentially this boils down to using surface sytactic constraints to hint at underlying semantic constraints. fascinating.

i wonder if verbs can be classified as those which can undergo VP ellipsis and those which cannot. for instance,

digging in and digging out

we do not say

digging in and out

to convey the same meaning. i wonder if this restriction is based on an idiomatic or metaphoric use of the verbs. i shall have to explore this.