Mahler: Symphony No. 9. Vienna Philharmonic Orchestra, Bruno Walter. EMI Editions
Recorded 15 and (live) 16 January 1938 at the Musikvereinsaal, Vienna
EMI Références / Great Recordings of the Century CDH 7 930292 2 (1989), barcode 077776302928 (transfer by Keith Hardwick):
EMI Great Artists of the Century 5 62964 2 (2004), barcode 724356296428 / 5 62965 2 (2004) barcode 724356296527 (digital remastering by Andrew Walter):
Note : the Angel logo on the bottom right corner of the front cover makes me think that, although “made in Holland”, 5 62965 2 was the US edition and 5 62964 2 the European edition
Also compiled in “Bruno Walter The Early Recordings”, 9 CDs EMI / Warner Icon 6 79026 2 (2012) barcode 5099967902620:
Despite the 1938 sonics (much improved on the 2004 remastering), a musical experience unparalleled to this day
Originally posted on Amazon.com, 14 November 2011, reposted here 1 April 2021
One can dispute that it is always great recordings of the century that EMI has reissued in its “Great Recordings of the Century” or “Great Artists of the Century” collections, but Bruno Walter’s conducting of Mahler’s 9th Symphony, recorded live in Vienna on January 13, 1938, certainly is. First, because it is the premiere recording of the symphony, Mahler’s penultimate (that’s counting the unfinished 10th), first sketched during the summer of 1908 as the composer was working on Das Lied von der Erde, then completed in draft form in the summer of 1909. Second, because Walter has unique legitimacy in this work. He was then Mahler’s favorite disciple and there were long exchanges of correspondence between them in that period, although Henry-Louis de la Grange, in his mammoth Mahler biography, doesn’t record that Walter discussed the score with the composer as he did with Das Lied. Finally, Walter premiered the piece on June 26, 1912, shortly after Mahler’s death, and already with the Vienna Philharmonic. I hesitate to call Walter the closest recipient of Mahler’s intentions as can be and his most truthful interpreter, not only because this recording dates from more than a quarter century after the premiere and almost thirty years after the work’s completion and any possible conversation Walter might have had with Mahler about it, but also because Mahler himself considered that the composer’s intentions were never definitive and were only those expressed on the day of performance. So there can be no certainty that Walter’s interpretation in 1938 can give an idea of the way Mahler would have conducted it, had he not died. Still, there is a direct line between Walter and Mahler, that only Mengelberg – another early champion of Mahler’s cause, much appreciated by the composer – can emulate; more important still, Walter’s interpretation in 1938 is unique and special enough to substantiate at least the thought that this is as close as we’ve ever had to the way Mahler would have done it. It is far removed from any later interpretive paradigm in this piece, especially in its adoption of brisk tempi in the outer movements that (other than the maverick Scherchen in a recording made for the Austrian radio in 1950 but that was published only in 1990 on Orfeo C 228 901 B – link will open a new tab to my review), no subsequent version that I know has paralleled, and it is filled with a unique emotional intensity.
Certainly, amends must be made for the constricted 1938 sonics (integral with audience noises and coughs in the background), that prevent much of the enjoyment of Mahler’s score, with its myriad instrumental details and indications of articulation, dynamics, playing modes. In particular, Mahler’s scoring in the first movement calls for a distinct spatial separation between first and second violins and you don’t hear it. The horns also are most of the time barely audible, and likewise with the woodwinds, indistinct in the tutti (but clear in the solo passages). Mahler (Mahler’s compositional processes and, hence, the full enjoyment by the listener of Mahler’s compositional output) is not just about overall impression, melody over a vague background of sounds: it is about the intricate interweaving of multiple strands, and all these need to be heard; it is also about the different instruments’ specific timbres and character. Here, despite the major improvement brought by the 2004 remastering over the previous one from 1989 (with less swish noise and cracks from the 78rmp surfaces, more orchestral presence and much more impact given to the brass outbursts, but also a little more glare and harshness, and even a measure of saturation in some climaxes, such as at 19:20 in the first movement, which wasn’t there in the previous remastering), it is frustrating when, in the tutti, you hear mainly violins and trumpets-trombones-tuba against a mass of indistinct sound blurred in timpani rolls (try for instance the climax at 9:48 in the first movement, “Mit Wut” – furiously). And you will never get the instrumental presence and pungency of any modern recording.
But musically, Walter’s reading remains, still today, in a category of its own. Walter has a firm grasp over the first movement’s complex structure, his contrasting tempos are well chosen and tempo changes organic. By today’s standards his tempos are fast (compare his first movement’s 24:40 to the 29:14 of his 1961 remake, to Horenstein’s 29:12 in 1952 on Vox (VoxBox Legends CDX2 5509), to Bernstein‘s 28:15 in 1965 or to Levine‘s 29:32 in 1979 on RCA – and I haven’t specifically looked for the longest ever) but most of the time it is not something I perceive when I listen to the recording but only something I know by looking at the clock. What I do perceive is the fiery intensity of the approach, which develops as soon as the first forte, at 2:03 (measure 29), introduced by a crashing brass chord (so much more impactful too in the remastering). Walter doesn’t drag and is even pressing in passages in which modern interpretation tends to drag, as at the return of Tempo I at 6:26, measure 110, but with firm grounding in Mahler’s instruction, here, not to drag (“nicht schleppen”). His subsequent “Mit Wut. Allegro Risoluto” (9:48) is indeed resolute and furious, and the ensuing “leidenschaftlich” (passionately) at 11:08 is again furiously passionate. Only in the “Schattenhaft” section (“shadowy” – what a great character indication) at 13:24 (measure 254) did I find that there could be more a sense of a stasis than Walter’s brisk tempo and feeling of urgency can convey. But Walter is not just pressing urgency, and the way he has the Philharmonic phrase the opening motto first played by the second violins is filled with the warmth so typical of him.
The Ländler-like second movement fares better sonically, with more of the intricate instrumental details coming through, and Walter captures remarkably well the music’s saucy moods (“täppisch”, “schwerfällig”: clumsy; “flott”: jaunty, lively, perky, “keck”: brassy, sassy), with great instrumental character and pungency (the oboe and clarinet’s “duck sounds” at 5:30 are irresistible, for instance). His held-back tempo at the start brings a feeling of good-natured bonhomie, but his brisk Tempo II at 2:53 (much brisker than Mahler’s “POCO più mosso” seems to imply) is truly savage. There’s no discussing an interpreter’s choice of a basic tempo, since Mahler gave no metronome marks and even his character indication at the begining is opened to much freedom of interpretation: “in the tempo of a comfortable Ländler” – a Ländler is defined by the Duden dictionary (in backward English translation) as “a slow folk dance in ¾ time”. Still, I find that Walter doesn’t always carefully observe the relations between Mahler’s three different tempos, and in particular that he doesn’t distinguish enough between the very slow (“ganz langsam”) Tempo III at 5:02 (measure 218) and the tempo I which returns “piu mosso subito” a few bars later (5:30): he grasped that tempo relationship better in his 1961 remake. Likewise, his “fliessend” yet a few bars later (6:25, measure 256) may not be flowing enough, and for sure, he doesn’t go back “a tempo (wie zuvor/as before)” after the big ritenuto at 8:21 (measure 346). These problems derive partly from the adoption of a Tempo I that is too slow in relation to the Tempo III, and for all its character I find that the alternative offered by Barbirolli in 1964 with the Berlin Philharmonic, and Bernstein in 1965, with a considerably brisker Tempo I, less good-natured and more exultant, allows a better realization of those tempo relationships. But those are details perceived only with score in hand, and Walter’s overall picture is entirely convincing. Interestingly, when Walter re-recorded the symphony in 1961 for Columbia with the Columbia Symphony Orchestra, he didn’t change anything significant to his earlier conception (see my review).
The fury of the Rondo-Burleske has rarely, if ever, been equalled (although the neglected Leopold Ludwig, in the first stereo recording of the symphony, made for Everest in 1959, came close. You wonder if the Vienna Philharmonic will have any reserve left when comes the double acceleration of the coda (Mahler’ whips up the already very fast movement to “più stretto”, than “presto” – they do. If you are listening carefully, some instrumental details will make you jump out of your seat like the furious first attack of the strings (0:02), or the forte trombone chord at 3:07/measure 196 ending on a sf, sounding here like a herd of elephants trumpeting.
Again, by today’s standards (of for that matter, by the standards established as early as the second studio recording of the piece, made by Jascha Horenstein’s in 1952 for Vox, see link above), Walter’s finale is very swift – it clocks at 18:04, to be compared (for instance) to himself in 1961 (21:02), to Horenstein’s 25:11, to Bernstein’s 22:55 in 1965 or to Levine’s extreme of 29:47 in 1979 – and on the face of it you might anticipate that speed came at the expense of depth or interiority. Not at all. Certainly, in the various “sehr langsam”, “molto adagio” and “sehr gehalten” (very held back) sections, there is scope for more of a sense of stasis and suspended time. But it is a tribute to Mahler’s genius that the emotions stirred are NOT a function of tempo, and Walter’s urgent approach again lends the finale a unique intensity. Moreover his forward-moving tempos allow him to make more sense of Mahler’s various indications like “fliessend” (flowing – everything’s relative of course, but what’s flowing when one moves from a standstill tempo to a stalled tempo?), “drängend” (“pressing”), “straffer” (“thighter”). Walter’s finale is not any kind of Buddhist accepting Abschied (farewell) to the world, it is a pained but passionate declaration of love of the world.
Being out of copyright in many countries this recording has been reissued on various other labels than EMI, too many to list them all (and not worth it when it comes to the bootleg Italian labels), and anyway I haven’t heard them to compare the value of the transfers. Still, notable other editions include Dutton CDEA5005 (1996) barcode 765387500528 (haven’t found suitable photos online), reissued on CDBP 9708 (2001) barcode 765387970826 (transfer Michael Dutton) and Naxos 8.110852 (2002) barcode 636943185229 (transfer Mark Obert-Thorn); the review published in The Gramophone is currently (March 2021) available online.
Among the other editions, Opus Kura is noted for its non-interventionist approach: little or no filtering and other tampering, which means preserved frequencies, but lots of surface noise: OPK2060 (2006) barcode 4582158680601; there is also a 2017 remastering, OPK2121 barcode 4582158681219
There is also a reissue by Yves Saint-Laurent Studio, YSL 095978, a label that has garnered much praise for the quality of its transfers: see entry on their website. Andrew Rose of Pristine Classical has also reissued it, PASC389, see entry on their website, with a review comparing the merits of Rose’s and Michael Dutton’s respective transfers.