This continues Gustav Mahler — Symphony No.10 (Part 1).
Twenty minutes into the Adagio of his Tenth Symphony, Mahler puts down two horrible, threatening nine-tone chords that remind of the ‘gate posts’ with which Beethoven’s Eroica opens. When heard in performance, they open that movement up for the first time: Here, after all, stands the terrible, black, and towering gate to modernity from which, once entered, there is no return. (A nine-tone chord might suggest that, Mahler was just three notes from having quite naturally—without any musico-ideological plan—stumbled upon a twelve-tone system of his own.) The descent from these gates tellingly brings us—via the amorphous Scherzo—to the “Purgatorio” middle movement. Mahler wrote into the margins: “Dem Teufel tanzt es mit mir”. (“The Devil is dancing [it] with me” – see picture accompanying Part 1.)
Where the travel goes from there, I cannot tell. There are moments in the finale, after passing those two horrible gates of chords again (this time some ten minutes into the movement), that suggest an ascent from the Purgatorio to a more solemn peace. The haunting flute melody (still before those gates), for example, which meanders with haunting beauty in all the performing versions, already hints at a contentedness otherwise akin only to the finale of the Ninth. But first that finale opens with seven, later repeated, dreadful (blunted) blows on…—well, and here the versions differ: a bass drum like Wheeler has it? A military drum as Carpenter suggests? Muted (as Cooke suggests) or forced, loud, and unfiltered (as Rattle departs from the Cooke score in both his versions)? And should there be two—one closing the fourth and one opening the fifth movement, as indicated by the score? Or one, to connect the two movements, as does Kurt Sanderling in his seminal 1979 recording does and as Simon Rattle and Daniel Harding have since adopted, too. These blows were apparently inspired by the passing of a cortége for a Fire Chief below Mahler’s hotel window in New York. It makes sense to hear the blows as the sound that traveled up to Mahler, possibly emitted from short ‘tattoos’ on a snare drum. If so, the vigorous single strokes of Sanderling (or Rattle I, slightly less glaring but even louder, Hammerblow-like) make less sense, whereas the washed sound of a distant drum (hence hearing the ‘tattoo’ as one solid thump) does. I have heard none more convincing than the one Barshai employs at this crucial moment in the score. Noseda opts for loud but not harsh, Harding for soft, but with too much reverb. Inbal and Gielen could both benefit from a more muffled sound.
Since the Tenth only exists in ‘performing version’, there is no point getting too deeply into the different approaches; unlike with the Sixth’s two contentious choices regarding the Hammerblow-count or the movement order, the differences in the Tenth are the point of the various versions and part of its reception. Suggesting that one touch or another works better—like Cooke’s use of brass (tuba) in the opening of the last movement, which distinctly sounds like Wagner, whereas Wheeler, Mazzetti, and Barshai, using double basses, get a more discernibly Mahlerian sound—should only suggest that we are lucky to get more and more versions of this work, all of which work off each other and can utilize good ideas and melt them with new ones. Anything else would be precisely the kind of presumptuousness that has made all the re-constructors approach this task so tepidly in the first place.
Since its appearance, Rattle’s Berlin recording of the third Cooke edition (with some of his own touches already displayed in the earlier Birmingham recording), has been hailed as the go-to ‘performing version’ tenth. Glorious sound and what was at the time deemed the most up-to-date edition as well as the fact that Rattle has been (and remains) the foremost champion of the ‘complete’ Tenth, have largely contributed to this. Since then, there have been a few other Cooke III recordings, though, and none of them worse and some better than Rattle: Michael Gielen’s on Hänssler (the most successful, most unforgiving, most terrifyingly modern), Daniel Harding’s on DG (superbly played by the Vienna Philharmonic), and Gianandrea Noseda’s on Chandos (more unified than Harding, a tremendous finale with a haunting portamento from ppp to pp, albeit with fewer ‘moments’ elswhere). In terms of sheer beauty, none of these recent releases manage to match Riccardo Chailly and the RSO Berlin (Cooke II). Interpretively reserved, but shaped with loving dedication. Inbal (Frankfurt RSO, Denon, oop) always keeps my interest, but Sanderling (Berlin SO, Berlin Classics) no longer sounds as good next to the newly appearing competition. The playing of the Berlin Symphony Orchestra is not at an exalted level and he rarely gets the pianissimos to be truly hushed. It is interesting to hear his stylistic influences and textual choices and deviations on Rattle, but all the above named sound better and the orchestras play better, now that the Tenth is no longer a nose-wrinkling experiment to them.
Given the improvements (notably some trimming-down) of Cooke II and III over Cooke I, the re-issued Ormandy is an interesting piece of history, and an engaged performance, but not a viable choice in light of many other good performances of Cooke II or different versions. Wheeler has gotten its first commercial recording on Naxos—with Robert Olson (he also led the premiere performance) who directs the Polish National Radio Symphony Orchestra in a fine, if not particularly polished performance. The result may have warts and the minimalist approach to Mahler's sketches may take some getting used to, but the version and the recording are better than its reputation. Not a first choice, but a first second choice. The spartan Wheeler/Olson version turned out to be so good—so close to Mahler’s draft—that Mazzetti went back to revise his own version (originally recorded with Leonard Slatkin and the St.Louis SO), which in turn was played and recorded by Jesus López-Cobos and the Cincinnati SO. Leonard Slatkin suggested that the recorded first version should be very similar to the second version, since the latter version incorporates all the changes that he suggested to Mazzetti before performing his first completion. (In conversation a few years ago, he even suggested that he might like to try his hand at completing a performing version himself, if he finds the time.) But as Remo Mazzetti points out, the changes that Slatkin unilaterally imposed on the score (substituting, for example, the xylophone for the glockspiel in the last 2 movements, much to Mazzetti’s dismay), have nothing to do with his—Mazzetti’s—subsequent version... and were in any case not "suggested" but presented to Mazzetti as a fait accompli. A truly idiomatic recording of the non-bowdlerized Mazetti-solution has yet to be issued, but may well be (or at least should be) forthcoming one day.
Carpenter, who worked on his fairly liberal version without knowing of the parallel efforts by Cooke, seemed less inhibited in adding and inventing to the Mahler facsimile, and perhaps he overshoot here and there. The result is, together with Barshai (but less happily, to my ears) at the other end on the re-constructive continuum as measured from Wheeler. The recording to go with Carpenter is Andrew Litton’s with the Dallas SO. The Dallas Mahler recordings, even if I have not mentioned them in the past survey, are good; this Tenth (for obvious reasons) being the most interesting and the Second the most impressive of the lot.
Rudolf Barshai set upon his own draft, drawing on all available versions and coming up with what I find to be by far the best attempt of presenting a coherent, exciting Mahler symphony. Excellent sound and a wildly inspired playing Junge Deutsche Philharmonie add tremendously. Barshai uses an almost bewildering variety of instruments in his Tenth, thereby moving further from the score in that regard than anyone else. Some might find the sounds of the guitar, castanets, or xylophones as uncharacteristic of Mahler (the Mahler of Symphonies 1-9 that is) or could think the atmosphere ‘congested’. Maybe, but it makes for tremendous excitement. Of course we have no idea what Mahler would have ended up using for the final version of the Tenth—and despite the curiously large number of percussion instruments that Barshai uses, the tender and sparse, ‘broken’ orchestral texture of the symphony never gets disturbed. No one else sets the two gates in the first and last movement down in such a deliciously terrifying manner; Barshai successfully circumnavigates those rare moments where Cooke sounds oddly un-Mahlerian or too literal. The additional meat he hangs on the bones of the Mahler skeleton—as compared with Wheeler’s leaner attempt—make for generally more satisfactory listening.
Overview of the whole Mahler Survey on ionarts at this link.
Mahler 10 Choices
1. Rudolf Barshai, Version: Barshai, Junge Deutsche Philharmonie, Brilliant
2. Michael Gielen, Version: Cooke III, SWRSO, Hänssler
3. Robert Olson, Version: Wheeler, Polish NRSO, Naxos
4. Riccardo Chailly, Version: Cooke II, Berlin RSO, Decca/ArkivCD
Also: Gianandrea Noseda, BBC Philharmonic, Chandos or Daniel Harding, Wiener Philharmoniker, DG (both Cooke III)