Blogging Pachelbel #3 -- Paillard: [Editorial note: when I wrote this I was under the impression that the recording I was listening to was the original 1968 version, but I've since found out that I was using the 1989 re-recording, which is shorter by a whole minute. When I get access to the 1968 version, I'll replace this entry with my review of it and add an additional entry about the 1989 recording, in the chronologically appropriate location on the blog.]
There is actually a lot to like about this recording, to be honest. After a rocky start balance-wise (the introduction of the 3rd couplet is almost inaudible) and a momentary abandonment of the canonic texture in the middle (guess where?), the sound of the strings is quite clean, with a relatively cool (perhaps French?) approach to the piece. The tempo is quite steady throughout, tending to sag a bit below the 37BPM average at the beginning, but regaining that later on.
But of course, there has to be mucking about with the music in the middle, just as in the other two recordings considered so far. In this case, no measures are dropped, but the two couplets of the canon at mm. 27-30 are never audible. Instead, at m. 27, a completely unwarranted harpsichord solo begins. It does not replicate the figuration of the canonic part, nor is it canonic. It's just four measures of repetitive noodling about before the strings creep back in at m. 31. The harpsichord continues playing its noodling as the lower parts re-enter following the first violin, and when the 3rd violin enters with the repeated-note couplet, the harpsichord basically vanishes back into the background.
Why?
Well, there was clearly something about either the passage as originally written, or about the perceived shape of the piece that those who planned all three of these performances saw in common. My guess is that after the increased activity of the 32nd notes starting in m. 19 and the gradual withdrawal of that rhythmically active melody ending with m. 26, that the next few couplets were somehow seen as too much activity before the start of the big crescendo to the end.
But it seems to me to be entirely the case that this idea of two parts with a dynamic arch shape in each is artificially imposed on the piece. It seems to me that the 32nd-note passage should be neither as loud nor as active as these recordings make it, and that if that is the case, the radically softer passage in the middle no longer needs to be shoehorned into the piece.
Worst of all for me here is that the music that Paillard and his cronies come up with here to replace what Pachelbel wrote is BORING -- the harpsichord figuration is just repeated over and over, with nothing interesting going on in it at all. Now, it is possible that the actual violin parts are being played very softly by one or two players, but if they are, the dynamics are so soft or the balance of the recording so drastic that the music is just not hearable. But it's also completely possible to make the 16th-note passages that follow the 32nd-note passages soft and understated, as is beautifully shown in a video performance by Voices of Music (San Francisco). There is nothing artificial about the way Voices of Music players handle this passage that seems to have given so much difficulty to Baumgartner, Muenchinger and Paillard. Perhaps it's because they were using an orchestra, but in that case, judicious use of solo/tutti might have gotten the job done. Or maybe our present-day players are just a lot better at the style than they were back then.
But jumping ahead to another recording, I have found a performance that plays the passages that these recordings excise -- 101 Strings plays the music as written, with no changes except for adding a harmonization to the initial continuo statement and, perhaps a viola part (it's not particularly audible if it's there). But the treatment of the passage from m. 27 to m. 34 perhaps suggests why these earlier performers cut it or drastically altered its content. The 101 Strings whacks every 16th note in this passage as if it were pesante. Compare that to the treatment in the Voices of Music video, where each 16th note is bowed, but there is a variety of articulation from slightly detached to light portato to legato. It seems to me that modern string players have historically depended on taking multiple notes in a phrase under a single bow as their only method of playing legato. Whenever they see unslurred notes, they see one bow per note and they seem to me to interpret this as NON-LEGATO. Now, that may be an appropriate articulation in some passages, but in others, legato is more musical, i.e., groups of notes that are connected (though not under the same bow). To me, it's quite obvious that there are breaks at the leaps and the notes within each range should be smoothly connected (even though each is taken under its own bow). But in the 101 Strings recording, these notes are aggressively detached and heavily weighted.
It's ugly. It's unpleasant. And if it's the only way musicians of a certain era could conceive of realizing music notated in a particular way, then I can see why they thought something needed to be fixed. I'd much rather hear the Baumgartner or Muenchinger or Paillard versions than the 101 Strings travesty (and that's not the only recording that exhibits this defect).
After all that, there are two other aspects about this recording that are quite different from the two previously discussed. The Baumgartner pretty straightforwardly apportioned the three string lines between three violins, with a fairly understated viola part that is never obvious (I'm not even entirely certain there is one). The Muenchinger/Stuttgart recording adds some octave doublings here and there, but doesn't muck around much with the disposition of the original parts.
This recording quite famously adds the pizzicato viola line at the beginning. It's sappy but effective -- not at all inappropriate for an orchestral transcription, in my opinion. In a performance for the intended forces, a chamber group, you might very well have a continuo group with a plucked instrument like a theorbo that might fill in some figuration in just such a manner (though I'd expect our present-day theorbists to be substantially more imaginative in their figuration and rhythmic subtleties).
The other unique aspect of this recording in comparison to those that I discussed previously is that it extensively re-orchestrates the ending, adding octave doublings both above and below. Specifically, in the couplet introduced by the first violin in m. 49, the violins play with a double octave above the written notes. This makes for a spectacular splash of sound when they go up to the high D, and it only intensifies as the other parts get in on the act. Doublings an octave lower (presumably in the violas) are also heard in the octave leap couplet comprising the last 4 measures of the canon.
As orchestration goes, these additions are fairly effective, but I particularly see the added 8va passage as guilding a lily that was already quite stunning. But not even 101 Strings does this, and splashy string sound was supposed to be their trademark!
