The Wagner Cube
I've had a chance to digest those parts of the behemoth Decca set, Wagner: The Great Operas [sic] from the Bayreuth Festival, with which I didn't already have some acquaintance. That is to say that the Böhm Ring and Tristan (both from the late Sixties) have been in my collection for some time now, so I didn't feel the need to listen to those music-dramas one more time. Indeed, for the latter music-drama (the importance of which to Wagner's oeuvre can hardly be overstated), I generally recommend Leonard Bernstein's SOBR 1983 recording on Philips. It's not without its flaws, but Bernstein knows that you can't hurry love - or Tristan, depending on how metaphysical you want to get.
Bernstein's Tristan (like Donald Runnicles' BBCSO recording of the same on Warner) was done, as I recall, over a series of live concerts done one act at a time. It is not, however, from Bayreuth, which means that it is still as obscure today as it was before the Cube dropped. It seems that the Universal people more or less released the bulk of the Philips' Richard Wagner Edition, with some key changes (Sawallisch's 1961 Holländer for Nelsson's 1985 recording and Böhm's Ring for Boulez', which isn't as big of a transition as the other one). That means, of course, that if you like the approach to Wagner of either Wolfgang Sawallisch or Karl Böhm, then you're in luck. You'll get the former's Holländer, Tannhäuser, and Lohengrin - all from 1961-62. What you'll get from the latter I mentioned earlier.
The Sawallisch stuff is nice, since it documents Bayreuth at the end of the Golden Age. There are better documents for Höllander and Lohengrin (Keilberth '55 and Von Matačić '59, respectively), but compared to what was coming down the pike, there's not much wrong with Sawallisch's entries. His conducting is pretty straightforward by Wagnerian standards: he is certainly in the Joseph Keilberth mold of conductors (Sawallisch followed Keilberth in Munich). I suppose you could, in a fit of derogation, call Sawallisch a Kapellmeister, but that would imply that he spent years working through the same material (though he did début very young at Bayreuth). Nothing wrong with that, given the proliferation of conductors who really have no business touching a Wagner score with anything other than a wistful "Someday..." on their lips.
His readings are all very solid, musically speaking, even if I prefer other conductors and other singers in the same material. As fits the time when these were recorded, Jess Thomas and Wolfgang Windgassen are his tenors (Lohengrin and Tannhäuser, respectively) and Franz Crass is his Holländer. Really, I have always felt that Thomas was at his highest and best use as Parsifal for Hans Knappertsbusch in 1962, though I prefer Jon Vickers from 1964, despite the mono sound. Anja Silja features prominently as Sawallisch's go-to gal for the soprano roles of Elsa, Elisabeth, and Senta. I don't know that I am in love with Silja, but she's not terrible. I could go on about these recordings, but they're well-known and known quantities just like Karl Böhm's Ring and Tristan (love 'em or hate 'em - you know where you stand).
What were, however, less well-known to me were, of course, Silvio Varviso's 1974 Meistersinger and James Levine's 1985 Parsifal. I knew the latter somewhat better than the former through a highlights disc I had at one point, but one does not "know" any Wagner music-drama, much less Parsifal, through a ten-or-thirteen-track highlight disc. I'll start with Levine's work, though. This is a slow, slow Parsifal. Of the nine versions I have, this one takes the prize for the longest runtime. Here's some statistical information that might put that into context:
Barenboim '91 ---- 4:16:12
Knappertsbusch '64 (Orfeo) ---- 4:15:46
Karajan '80 ---- 4:15:09
Kubelík '80 ---- 4:12:44
Knappertsbusch '62 ---- 4:09:53
Thielemann '06 ---- 4:02:11
Kegel '75 ---- 3:40:06
Boulez '70 ---- 3:38:34
Now, runtime doesn't say a whole lot, since a good conductor is going to go at a pretty good clip in some places and linger in others, but you can begin to see a correlation, at least from the orchestral contributions, between versions that are pretty solid and a runtime around ten minutes (rough, eyeballed average) over four hours, with Thielemann at the low end and Barenboim at the high end. Levine takes almost thirty minutes more than that, or, to compare it to another, equally radical version, about a hour more than Boulez did in 1970.
Flexibility be damned, a recording with a runtime like that is going to be a little slow. Levine's Parsifal is slow. That's not necessarily a bad thing, but Levine's recording falls short of the standards set by Knappertsbusch and Kubelík. A spin of the recordings by those two conductors shows where the problem is. They let the music dictate the terms of the performance. Their approaches are tailored to allow the internal pulse of Wagner's score carry through and carry the day (you might call it a sensitivity to the longest lines). Levine's Parsifal, though, is a little bit stretched. Call it the opposite of the Boulez set, which is Parsifal compressed and thinned out to reveal the interior architecture (which really isn't idiomatic Wagner); in this case, it's Parsifal expanded to emphasize what would have been clear from the music itself.
Is it a bad performance? Well, no. It's probably better to add fifteen or twenty minutes and have a blogger twenty-some years later call it a slightly self-consciously grand and affected performance than it is to cut half and hour and strip away Wagner's sound. Knowing both the Boulez and Levine sets reveals the brilliance of Hans Knappertsbusch and Rafael Kubelík because it shows how precarious the balance of a great Parsifal really is. Take it too fast, and it sounds weird; take it too slow and it feels a little affected in places
This post has gone on far too long, so I'll deal with the Varviso Meistersinger later.