...wicked
I'm a little confused. Unless you have access to have the music written in a way to naturally leave room for narration vox there really isn't any way to practically do what you're asking without lowering the music volume (via whatever technique works). I'd reconsider your premise that the music always needs to be out front in relation to your narration unless you have a wacky plan for one or the other (doing an experimental treatment on the narration vox for example).
 

It's something to think about.
Let's take a classic example: Tom & Jerry. The music plays practically all the time, and eventually there is some narration.
You mean in this case that music should always be limited to, e.g. -10db? And the voice would be above, e.g. 0dB?

I don't believe anyway that there is any effective automatic solution.  An automatic solution is always going to sound automatic. eg dialogue comes in, music drops (jumps) down, dialogue goes away music rises (jumps) up to a similar level to dialogue etc. They have automatic solutions on radio but it seems to work there.
 
If you are really serious about this, manual automation is really the only way to go.  What is the problem in doing that.  Are you so pushed for time and the budget so low that proper automation is not possible.  If that were the case I would not even take the job.  Time and money must be allowed for the mixing of the soundtrack.
 
It just sounds so much nicer when this is done properly.  As I have said music can ease down under dialogue and ease or rise so nicely you hardly notice the effect etc.. All the great TV soundtracks you hear have been properly mixed.  Time has been spent in this area.  It is hard work and takes time but the results will show.  You end up with this seamless soundtrack that just flows nicely.  Sometimes it is nice for the music to sit just under the dialogue especially if the music is not too busy at that point, other times it needs to sit a little lower in relation to the dialogue.  These types of decisions cannot be made automatically, only you with your ears can do that.  Same thing when the dialogue stops, sometimes the music might not reach full soundtrack level because it sounds better 3 dB below that. Other times it is nice for the music to ease up to the full soundtrack level.  Sometimes certain vocal words or phrases also need to automated a certain way in order for them to achieve better clarity too.  Something Vocal Rider and things cannot handle that well.
 
Automatic systems and just not smart enough to make those sort of decisions.  I have sat in many soundtrack mixing sessions.  It is like a fine tooth comb over every second of the final soundtrack. That is why it usually takes ten times as long as the program itself in order to do it.  I have seen some soundtrack mixers spend over an hour getting 5 seconds of a final soundtrack mix right.
 
And to your last post too, no, leaving the dialogue at full level and the music to a certain level below that sounds terrible. As soon as the dialogue stops there will be a big hole in the audio eg the soundtrack goes quiet. Not good.  Often a lot of cheap adverts are mixed that way and it sounds terrible.  Listen to high end adverts and listen to how good it all sounds in the end.
 
Listening is the key as well.  Listen to lots of TV soundtracks and you will start to get the idea.
 
 
 

rogeriodec
It's something to think about.
Let's take a classic example: Tom & Jerry. The music plays practically all the time, and eventually there is some narration.
You mean in this case that music should always be limited to, e.g. -10db? And the voice would be above, e.g. 0dB?

Uh, Tom & Jerry had narration? I thought all their sounds were generated by orchestral elements (except when they were actually "singing"). At least the ones that were out when I was young. I'll have to go and re-watch some!