What I do is first create an Aux track (mono or stereo), and put on it whatever EQ and Compression plugins I want to hear on the vocal (if any). I will avoid plugings inducing to much latency like Waves C1. I will also set up a few sends ready to go to whatever reverbs are hooked on the system.
The input of the Aux track will be a free bus (let's say bus 9)
Then I create a few audio tracks labeled in a logical way (V1,V2 etc) with all outputs to bus 9
All levels stay at unity on those tracks. Any level change is done on the Aux track.
That way, whatever track you use to record you *always* get the same sound, and the same level (Singers love that). Need a new track in 10 seconds ? just create one, name it, assign the output to bus 9 : done.
Once you have several tracks, you can leave them all open but you mute the parts you don't want by selecting, separating and Command-M for muting/unmuting.
After that it's up to you to decide whether you want to make a bounce to get one full track, or move the good regions one track or just leave things as they are.
I suppose there are other methods but that one works fine for me. :)