Separation of voices

[gouldfischer]

It is a given that Glenn Gould used to give great clarity to the counterpoint voices. This is certainly true, in the sense that the listener can easily follow the horizontal lines with ease. No confusion at all! No need to forget the melodies and just sense the verticality here and there! Everything is very well explained.

But...

I have a serious complaint about his Goldberg Variation #3 (and this applies to BOTH his studio recordings!!!). He plays it in such a way that it seems we have voices divided by pitch (treble/bass, soprano/alto) in the RH, instead of only ONE MELODY being repeated one measure behind. I mean, there is no tone, touch, volume or color differentiation between any two consecutive similar phrases, as if voice 1 always sang the treblemost and voice 2 the bassmost notes.

Let me use a visual abstraction now. IMHO, the "correct", in such a "canone all'unisuono", would be to paint the main melody with color 1, let's say, blue (starting at measure 1) and the 1-measure-delayed repetition (starting at measure 2) with color 2, say orange. This way, everything which is painted blue is repeated, 1 measure later, painted orange.

Instead, what Gould does is (as I perceive it): he paints some phrases blue TWICE (in the main line AND in the canon) and some other phrases red TWICE as well.

Another thing is that, sometimes, he seems to break the horizontal line (as if he were asking the OTHER VOICE to sing some notes from the ONE VOICE).

Look, for example, measure 7: the main line reads:
D... E F# G A B... D... C#... D.............................

Gould plays:
D... E F# G A B............... (one voice)
      (...)                D... C#... D........................... (other voice)


What do you think?

Cheers,
    Vinicius.

[rimv2]

 

uh


Whatever works... ah guess

[alzado]

I note that in the Myra Hess transcription of "Jesu, Joy of Man's Desiring," she continues to label the various parts by the singers.  For example, one section is named "Tenore" -- tenor.  [Note -- this is a piano solo! ]

I have never figured out if she expects persons playing the piece to try to somehow try to emulate the "Tenore" vocal qualities.

She didn't originally write this transcription for publication.  It was only published late in her life at the pleading of many of her admirers who loved it.  So I suppose it is hard to second-guess what she was doing.

[gouldfischer]

What I meant to say in that color analogy was:

If we were hearing a transcription of the Goldberg for, say, violin, viola and cello, I would expect to hear (besides the left hand being played by the cello, of course) the main melody of that Variation #3 being played entirely by the violin, and its echo, which starts in the second measure, entirely by the viola. Or vice-versa, it doesn't matter. It could be that the viola started the main line in the first measure and then the violin echoed it. (I've actually heard such a transcription once, by Dmitry Sitkovetsky.)

What I would NOT like to hear, in this piece (it is a canon, after all) would be like the violin playing all the RH's treblemost notes whereas the viola played the RH's bass ones all the time.

To make such a separation of "colors" in the piano is not that easy, but can be done. (Well, I suppose I myself can achieve that somehow in this very piece.) Of course I'm not saying that Gould was not capable of doing it a thousand times better than I do, I just wonder why he did not WANT it to be this way. Perhaps he intended a different effect: that of an IDENTICAL echo, not as if played by a different singer or instrument following the leader, but rather as if we recorded the leader and them played it back twice with a 1-measure delay. This is also tricky to accomplish in the piano, because the fingering is often completely different on each time a certain phrase is played (depending on whether it belongs to the leader or the echo).

Anyway, I guess it's a good practice to learn how to make it satisfactory in both ways. Perhaps I'll use the ritornello for that (the first time distinguishing the two voices, by means of a slight tone differentiation; and the second time making it so that the echo is a perfect copy of the main line).

What do you think about this? Is there a "correct" approach to this kind of canons?

Tx,
    Vinicius.