A good sound engineer should be able to handle any job and achieve a result that is acceptable to the audience. If he can't, well he shouldn't be there and the performance should be given purely acoustically.
There is often a requirement for a narrator to be present. This could be a work like 'Peter and the Wolf' by Prokofiev. Or it could be a children's concert where a presenter introduces the pieces.
Either way, the narrator will have to be amplified. Of course, the orchestra can be heard perfectly well just as it is.
The problem here however is that if the narrator is in front of the orchestra, then his or her microphone will pick up the orchestra as well. And that sound will be amplified too.
When the audience can hear both the natural sound of the orchestra, and the amplified sound from the speakers, the amplified sound is completely different in character and sounds pretty dreadful.
This will be most easily noticeable when the narrator isn't speaking. Ideally therefore the narrator's fader should be pulled down between narrated sections. Some operators prefer to cut the level by 20 dB or so, so the channel isn't completely muted.
This will solve the problem, but raises two other issues.
Firstly, if there is no script, then it will be impossible to know when the narrator is going to start speaking again. Plainly, it is preferable to have a script, but this doesn't always happen.
Secondly, the narrator might speak while the orchestra is playing. If you pull down the fader when the narrator isn't speaking, you will draw the audience's attention to the fact that sometimes the orchestra is amplified, sometimes it isn't.
The solution in these two cases is to position the mic as close to the narrator's mouth as possible, while guarding against pops and breath noises. This way, the orchestra will be quiet in comparison.
Often a miniature microphone will be used. The best place for this is on a short boom projecting from the ear, which is conventional theater sound practice these days. However, the narrator is now free to turn (which they wouldn't be with a stand-mounted mic), and as they turn towards and away from the orchestra, the amount of orchestral pickup will change.
With a mic mounted on a boom, this should not be audible. If the mic were clipped to the chest area, then potentially it could be.
The other possibility is to place a clear perspex (plexiglas) screen behind the narrator, so the mic is shielded from the orchestra. This might seem like overkill, but it is commonly done when a drum set, guitar and bass guitar are incorporated into the orchestra.
It has to be said that in practice, this problem is often not properly attended to, and the sound isn't as good as it could be. But a little care and attention can go a long way, and sound engineering should be about achieving as high a standard within whatever the timescale and budget allows.