The historical justification for the use of one voice per part in Bach's choral works remains a minority view even amongst period/historically-informed performers. What are anyone's thoughts on this practice, how do they weigh the historical evidence, do they think it is appropriate in this day and age, how do they perceive the musical results, and so on?

(You asked for this, Ollie! )

Musically I'm with Ollie here. IIRC one of the pieces of evidence in favour of OVPP is a painting that turned up in the 1990s (my recollection is very sketchy here) showing Bach conducting a four-voice choir. Anyway, we had an impromptu run through a couple of cantatas with minimal forces one afternoon at Nottingham Uni and the results to me were an absolute revelation. Just seemed to make absolute musical sense, no matter what the historical evidence (or otherwise).

On a slightly related note, I once took part in a catholic church service on Trinity Sunday.  The choir was only 7 strong, so to play a triple choir Gabrielli mass they had booked five trombonists!  Once we (er, I) had supressed the urge to tonk it was surprisingly easy to balance and sounded rather fine.

So back to Bach it shouldn't be too hard to balance a vocal quartet with a very small orchestra.

Isn't there a dilemma with the b minor mass of how to balance the sopranos, or is it 5 parts throughout (even in the chorales)?

NB

Oops.  I did realise yesterday, Mass != Cantata.

I'm afraid my main source of Bach knowledge is the 371, which offer tantalising glimpses of obligato parts in various instruments (2nd horn?!)

NB

Parrrott's book has quite a bit of useful information on this. Bach's manuscript is headed Gott ist mein König âb 18. è se piace 22. In other words although he's written for trumpets and drums the ripieno singers are seen as being (a) optional and (b) only an additional 1 per part.

Yes, ...trj... , I remember that run-through! Following which, we performed quite a few pieces by Bach with OVPP in Nottingham over the years, including some more cantatas, the Ascension Oratorio, Magnificat, parts of the Christmas Oratorio and B minor Mass ...

I'm with you and ollie on this, and for much the same reasons. The evidence presented by Parrott is very compelling, though opponents of the idea have pointed out that it's not 100% conclusive, whilst neatly forgetting that most evidence (including their own for the contrary arguments) very rarely is.

Contemporary pictures suggest that instruments usually outnumbered voices quite heavily. Some have argued that this is becase instruments are more interesting (and less difficult?) to portray than singers. But the surviving musical materials seem to suggest similar ratios. Coincidence? Did someone really go through all the voice parts, methodically destroying duplicate material, but forget to do so with the stringed instruments?

I think the clinching factor is that, to me, the result with OVPP, specifically in terms of the vocal lines in the "choruses", is so much more credible. I often used to hear singers complaining that Bach didn't know how to write for chorus: maybe it's because that wasn't what he was doing? The voice parts in most of his choruses are so virtuosic and soloistic that how anyone can conceive of them being for choir (even a chamber choir) is now a mystery to me.

I remember one very sniffy Radio 3 commentator mentioning that Rifkin was going to perform the Magnificat in the Thomaskirche with OVPP, and then adding, quite gratuitously and smugly, "Though how magnificent it will be remains to be seen." Even recently, on CD Review, the two presenters were strongly dismissive of the OVPP arguments.

In my experience -- I've directed performances of the Magnificat on separate occasions with chamber choir and OVPP, as well as playing in a few --- it's even more thrilling with just five singers and an appreciably larger ensemble! And the feeling for instrumental colour that comes through in even Bach's more modestly scored cantatas is so much more evident when you haven't got half a dozen (or more) sopranos belting out the chorale line at the top of their voices.

The thing is, the performing forces didn't all participate in the cantatas. There was plenty of music at the Thomasschule's various services (and other duties, including street singing of hymns I think) which wasn't remotely as complex.

One of the important things when interpreting the pictorial evidence is working out what they're actually singing. When singers are shown sharing from choirbooks for example I don't believe it's ever clearly a case of 'concerted' music.

Reiner, don't forget that, as well as voices breaking significantly later in those days (late teens, I think), these were "kids" who were receiving musical training (including sight-singing ability) from one of the most remarkable musicians in the history of Western art music. Even trebles from today's most elite choir schools would probably not come remotely close.

As for the argument that there were plenty of children singing in the choir -- both Parrott and Rifkin have argued persuasively that only a few took part in the concerted music: a team drawn from the squad, as it were. And evidence suggests that Bach will have used only the handful of best singers for his cantatas.

(This squad/team idea still applies today to most orchestras, for instance: you very rarely get the entire membership of an orchestra on stage at a single performance.)