Here's an article by Ton Koopman to mull over. I have to get my beauty sleep now.

I'm afraid that particular recording doesn't quite do it for me anyway...  I could point to a few reasons but the main one is simply that for me the vocal ensemble doesn't cohere. They're all (or nearly all ) fine singers but for me that doesn't quite add up to what's necessary. And Mark Padmore doesn't quite fit my idea of the Evangelist either... to be honest I find him a bit on the histrionic side (much more so than Dermota for Furtwängler for example!), especially as in the later stages of the story, which I know plenty of others like but for me it isn't the way to go.

In case I didn't say clearly enough that that was just my opinion I'll say it again.

IMESHO.

...although of course the singers didn't sing from a score but from individual parts.

Bother, I wanted to post the tenor part from the end of the Domine Deus and beginning of the Qui tollis in the Gloria from the B minor Mass but Photobucket is down for maintenance. I'll put it up when I get a chance. In any case the important bit is that it sails on from one to the other without so much as a double bar.

Indeed! And those individual parts would bear the word Tacet for those bits where singers were silent. If, however, one of the parts were to continue on its own, what would be the point of specifying (say) "Basso solo"? Surely it could only be to ensure that the part was to be sung by only one of the people reading the part.

Baz

But it is written against the vocal line - as also in Movt no. 8 where the solo Tenor is accompanied by all the violins in unison as well as by the continuo!

Baz

In any case, if the marking is found against the tenor and bass bits but not the soprano and alto ones it seems to me that the possibilities are the following:

- Bach was being completely methodical and meant the soprano and alto parts to be performed by ripienists as well;
- Or he wasn't and didn't.

In itself it doesn't change what anyone is suggesting about the vocal forces for the piece as a whole: even on the Rifkin/Parrott side of things it's well known that Bach often used an extra singer per part when trumpets and timpani were involved because sometimes ripieno parts have survived.

On the other hand those singers were used to reinforce a consort of single voices that sang the entire piece anyway - a distinction with the modern choir practice that's perhaps worth mentioning. I don't see that there can be any doubt with that point: Bach's parts for his 'soloists' also contain the corresponding lines in the choruses with no indication that they should stop singing.

A few more choice quotes -

Gottfried Scheibel, 1721:

"Even if tuttis... do occur, it is enough if the principal forces - even though they consist of single persons - do their job."

Mattheson, 1713:

Even "unpractised masters" should be able to play the trumpet with "such sweetness that even a recorder or some other gentle instrument may play in concert with it and be heard quite distinctly".

Scheibe, 1745:

"If there happen to be trumpets and timpani in the ensemble, one should hide them as much as possible and place them behind all the other instruments if one does not wish their rattling sounds to obscure harmony and melody as well as singers and instruments; in particular, one should always set them at a distance from the voices, as it is to these that they are the most detrimental."

Scheibe also recommends ripienists "where possible" where trumpets are involved. Note the "where possible" - even though his trumpets don't quite seem to have fitted Mattheson's earlier description...

On the other hand Mattheson is also the one who described the "howling symphony" of "the so-called chalumeaus" as something best heard in the evening from a distance "but never in January at a serenade on the water".

I like searching questions.