Mahler: Symphony No. 9. New Philharmonia Orchestra, Otto Klemperer 1967. EMI CMS 7 63277 2 (1989), EMI 3 8300 2 (2007) and other reissues

Mahler: Symphony No. 9. New Philharmonia Orchestra, Otto Klemperer 

Recorded 15-24 February 1967, Kingsway Hall, London

EMI CMS 7 63277 2 (with Wagner: Siegfried-Idyll) (1989), barcode 077776327723:




Reissue (remastered) The Klemperer Legacy  EMI 5 67036 2 (with Strauss: Metamorphosen, Wagner: Siegfried-Idyll) (1999), barcode 724356703629:




Reissue (remastered) EMI Great Recordings of the Century 3 8300 2 (with Strauss: Metamorphosen, Death & Transfiguration) (2007), barcode 094638000327:


Warner Classics (2016) barcode 0825646400751 (Japanese edition barcode 4943674227099):



Collected with Klemperer’s Mahler in EMI (France) 082265 2 (2011, 6 CDs), barcode 5099908336521:



Collected with Klemperer’s Mahler in EMI 2 48398 2 (2013, 6 CDs), barcode 5099924839822, new remasterings. Reissue Warner Classics with different logo, same cover and barcode. Warner also found with barcode 825646388080

MAHLER - Klemperer - Symphonie n°2 'Résurrection'








Klemperer’s radical, extreme Mahler: mass, deliberation and weight
Originally published on, 14 November 2011

The 1960s was a great decade for Mahler’s 9th symphony, as far as recordings went. First came Walter’s 1961 stereo remake for Columbia (he had made the premiere recording, a live concert in 1938 with the Vienna Philharmonic): an “old man’s” version perhaps (Walter would pass away a year later), underplaying Mahler’s shifts of tempo and keeping everything at a steady allegro moderato, but an old man still full of punch. Then, in 1964, came a passionate version by Barbirolli with the Berlin Philharmonic, and a coarse but even more passionate one (and one that remained neglected)  by Kirill Kondrashin on Melodiya. Bernstein‘s first go was recorded in December 1965 (but first released only at the end of 1967) and a fine version from Czechoslovakia came a year later, conducted by Karel Ancerl. 1967 saw the publication of the recordings of Klemperer and Solti (his first version, with the London Symphony Orchestra). Kubelik’s DG recording was also made that same year, but it is the one I’ve never heard; it has generally been considered understated. Finally the 1960s were closed by Haitink, with the Concertgebouw, and Abravanel with the Utah Symphony on Vanguard, both in 1969. The needs of the market were apparently satisfied for the next decade (or the labels thought so): the only versions of significance to be published in the 1970s were Giulini’s DG recording in Chicago (1976), then Levine’s with the Philadelphia Orchestra for RCA in 1979.

All those recordings were hailed by the critics (except the Kondrashin, which, although it was released in the US by Seraphim / Angel and received an excellent review in the April 1969 issue of High Fidelity, didn’t make much of a splash back then, possibly through limited circulation in the West,  and perhaps Abravanel’s, on account of his – or so it was deemed, but I don’t agree – lesser orchestra) and were (and are still) mentioned with accolades in the critical discographies However, listening again to Klemperer, I find that it can be endorsed only with strong provisos.

It may be a cliché to say that Klemperer, at least in his EMI years (but I’ve observed it also with some of his very early recordings from the 1920s) brought unique massiveness and deliberation to everything he tackled – but it is largely true, and it is verified again in this 9th symphony: his approach is massive, deliberate and granitic. In the first movement, all of Mahler’s promptings to accelerate, to play “fliessend” (flowing), “etwas drängend” (somewhat pressing), “allegro” (as at 6:18, measure 102) or even “furiously” (“mit Wut”, 10:38, measure 174), are underplayed or neglected in favor of a monolithic allegro moderato. Typical is the section starting at 16:16, where Mahler stages within seven bars a progressive acceleration of tempo (“nicht schleppen”: do not drag, “etwas fliessender”: somewhat more flowing, “etwas drängend”: somewhat pressing, “bewegter”: more agitated) leading to a lengthy agitated passage that eventually hurls into the movement’s second big crash at 18:18 (measure 314). Klemperer does lead a perceptible acceleration at “etwas fliessender” (16:30), but then from there he doesn’t go any faster, and on the contrary uses various high points to hold back tempo for sake of making expressive points (as at 17:11, measure 293), and he sure relishes Mahler’s “pesante” indications (17:40, measure 303). Finally, when comes the two and a half bars of “stringendo” at 18:14 leading to the great crash, the impression with Klemperer is pedestrian and heavy-footed, no race to the abyss but a shackled giant heavily falling to the ground. But his crash is blasting, almost (but not quite!) as much as Walter’s.

Commentators have tried to oppose Klemperer and Walter, based on Klemperer’s famous comment that Walter was a moralist and he an immoralist, but it seems to me, at least as far as concerns their respective recording of Mahler’s 9th, ill-founded. On the contrary they have much in common, and specifically the downplaying of the contrasts of tempo in the first movement, made up by the punch (Walter) and massive power (Klemperer). Whatever their differences, they shared one thing: they were both old men nearing death when they made their respective recording (Walter was 85 and Klemperer 81), and that strongly shaped their vision: not so much a case of the body not able any more to convey what the mind conceived, as of a mind turned unresponsive to these brusque Mahlerian shifts and surges of passion. For those, go to Barbirolli or Kondrashin.

Not that Klemperer’s version lacks tension. First because, even in the moments of ponderousness, like Walter he largely makes up in mass and dynamics what he looses in sheer drive. The “allegro” at 6:18 may be somewhat sluggish but it is certainly grand and powerful. His climaxes and crashes are overwhelming. There are also moments of great passion, as the “etwas fliessender” at 16:30, measure 279 (where, significantly, as mentioned above, Klemperer does speed up things) or again the surge starting at 20:20, measure 347 (although I whish that Klemp’ and his sound engineers had given more presence and intensity to the beautiful cello counter-melodies, measures 350 and after; they contribute a lot to the lyrical intensity).

Predictably, Klemperer’s spaciousness also generates much lyricism, starting with the great tenderness of the opening pages with their lyrical violin phrases that can assume many different and subtle emotional contents, from loving warmth to despair. There is also much atmosphere in all the typical funeral marches with tolling harp or timpani or double bass pizzicato and snarling brass, to which Klemperer, at a very held-back tempo, imparts a despondent ponderousness (and here I am using the word in a positive sense), a sense of burdened suffering. All those slow-motion brass fanfares (7:34 measure 125, 10:19 measure 168, 14:13 and after starting measure 243, 18:48 and after starting measure 323, where Mahler writes “wie ein schwerer Kondukt”, like a ponderous funeral procession) convey at Klemperer’s pace a unique mood of obscure, pent-up menace, death lurking over its prey and sarcastically observing its hopeless efforts to escape.

There is also a unique tension generated by Klemperer’s very radicalism, in his willingness to maintain a steady, unfliching beat, however slow it is. The best example is at the end of the first movement, with that floating, ethereal flute melody starting at 25:28. Unlike everybody before him, Klemperer keeps his steady, slow beat, never “cheating” here and there, never shortening the 4/4 bars by the value of a beat or a beat and a half, or shortening a silence from two beats to one. Before Klemperer, you might have thought that no flautist had enough breath to play exactly what Mahler wrote. The New Philharmonia’s flautist did – and when time came for Mahler’s “molto rit.” (26:15), he still did, and so did the oboist on his final, (very) long-held E.

Also noteworthy is Klemperer’s (and/or his sound engineer’s) great attention to the instrumental balances. The English horn comes out vividly at 1:03 instead of being left in the sonic shadow of horn and violins, and likewise, for the first time I can hear the clarinet at 0:35 and after (starting measure 9), where it is covered in all previous recordings by the main horn theme; in fact it is so clear that one also hears that all its phrases annoyingly end a breath after those of the horn, and I wish Klemperer and the session producers hadn’t let that pass quality control. In fact production is surprisingly careless in some in some details (apparently the recording was made prior to a concert, so maybe they didn’t have time for retakes). Klemp’ and his producers should have given the New Philharmonia’s solo horn a chance to make a retake of his duet with flute, to avoid setting in groove for eternity such glitches as those that happen starting at 22:57, measures 285, 286 and 290: not that it will bother most listeners, and it takes following with a score to really notice them, although the first one is quite blatant if you are listening carefully. But what you will hear though in the first movement even without a score (unless you are reading your newspaper or vacuuming while playing the disc) is how annoyingly out-of-sync pizzicatti double basses then cellos are with timps at 7:20 and 7:26 (measures 120 and 122). And I’m not sure what happened during the four seconds between 24:34 and 24:38, but harp and horn should have entered much sooner, after only an eighth-note rest; I think bass tuba and trombones failed to enter, leaving everybody awkwardly hanging.

Likewise, in the finale, at 12:33 into the beautiful trio for flute, oboe and clarinet (measures 93 & 94), the oboist (perhaps moved to distraction by the beauty of the music, unless it was by Klemperer’s sudden doubling of tempo in this section, see below) gets entirely out-of-kilter. In that same finale, the Gramophone reviewer, back in 1967, noted “a horn anticipating the first note of melody at the return of the molto adagio (on page 177 of the Universal miniature score)”: well, here, the wrong entry has apparently been excised. The reviewer also spotted one violin playing his next note a hair before the others (sure, it happens at 22:04 measure 169, 3rd beat, but it is really a very subtle glitch in ensemble playing), but failed to hear that the whole violin section enters a full measure too early at 23:17 (measure 180 – fortunately, they also enter where they should on the next bar). Other than that, the sonics are fine and notable for offering a clear separation of first and second violins, enabling you to fully hear their intricate counterpoint in the two outer movements.

Klemperer’s Ländler is exactly what you expect of him – truly hippopotamic, but, as in Walt Disney’s Fantasia, hippos wearing tutus. Just compare: Leopold Ludwig, the fastest before Klemp’ (and an unjustly neglected version, on Everest, see my review), got to the end of the first section (Tempo I) in 2:10; Walter in 1961, the slowest, in 3:04. Klemp? 3:13. It takes him 2:47 to get through the next section (Tempo II), against 2:27 for Walter and 2:22 for Ludwig. The movement’s TT is 18:35 for Klemperer, 17:30 for Walter and 14:06 for Ludwig. And to that Klemperer adds a few arbitrary touches, like the egregious stalling of tempo at 14:48 (measure 516). I dread the idea of anybody knowing the symphony and this movement in particular only through Klemperer’s recording: quite a distorted view it would give him/her! ! Imagine a kid getting a notion of classical ballet only through the hippos of Fantasia… That said, I enjoy Klemperer’s version immensely. It has a unique charm (Kung-Fu Panda as a toddler is another image that came to my mind), biting vigor despite the frozen tempos, and fabulous instrumental character, with more irony than insolence. Whatever you think of it, there is no questioning that it is absolutely unique.

The Rondo Burleske retains hardly more anchorage to normality. Massive and slow it is. Nobody here before Klemperer was beyond 13:30 (and very few after), he clocks at 15:13. But while this could have sounded only grotesquely ponderous, there is something implacable in Klemperer’s unique heaviness. It is not so much defiant (as Mahler instructs) and raging as massive and inexorable, “Big Brother” slowly but inexorably crushing the individual. You’d think a helter-skelter tempo (like the one adopted by Leopold Ludwig) would make it difficult for an orchestra to sustain precision of ensemble, but so does a very slow one apparently, and first trumpet threatens collapse at 1:47. But, thanks to the clear sonics, the slow tempo enables you really capture every detail of every instrumental line as in no other version before, and relish them.

After all this, it is almost surprising to see Klemperer returned to “normal” in the finale, which runs longer than most of its predecessors (only Horenstein in 1952, on Vox, clocked slower) only because its final pages, from 18:51 onwards, are taken at a really time-suspended pace: until then, it was well within average (which is to say rather flowing in comparison to the extremes of Levine in 1979 for RCA). Despite Klemperer’s curious choice to double the tempo at 12:04 (measure 88, the desolate section with woodwinds underpinned by tolling harp, where the distracted oboe adds an eighth-note), where Mahler instructs “stets sehr gehalten” (still, or always very held back), it is also beautiful, with great lyrical tension that sags only in a very few spots, and crowned by those magnificently restrained and sustained final pages, truly other-worldly.

I’ve compared the original 1989 CD reissue of Klemperer’s recording with EMI’s 2007 remastering on “Great Recordings of the Century”, and it would really be nitpicking to say that the latter has brought any significant improvement, or even difference – at least not to my ears, and not on my admittedly basic sound system (I’ve ripped both versions on my computer and used by Sennheiser headphones – sound’s good enough for me). It is also something I remarked with EMI’s 2002 remastering of Barbirolli’s Mahler 9th (see my review for a record of the various CD editions), and it raised the doubt that all this remastering gig may be only EMI’s scam to suck out more money from us. Anyway, unless you have a very high-end sound system, the deciding factor here needs not be sonics, but only price and couplings (the 2007 liner notes do give interesting anecdotes about Klempere’s understanding of the 9th, and how he had to arm-twist EMI in recording it). The two Strauss pieces are more generous than the earlier Siegfried-Ydill – but they could and can still be found in the original Klemperer series paired with Don Juan and Salomé’s Dance of the Seven Veils, EMI CDM 7 63550 2, barcode  077776335025. There is also another, 1999 remastering (they say), with yet another coupling re-shuffling (Metamorphosen and Siegfried-Idyll, see references in header), which I haven’t heard, and there are also a number of remastered Japanese editions on EMI-Toshiba, TOCE-3235-36 (1997) barcode 4988006730472, (HQ-CD) TOCE-90113/14 (2010) barcode 4988006877443, (SACD) Warner TOGE-15062 (2012) barcode 4988006898219 . The recent Warner reissue, following the “original jacket collectin” principle, is stingy. Your choice.







What matters is that Klemperer’s Mahler is, in its own way and on the opposite pole, as radical and extreme as Scherchen’s  – as examplified by the latter’s 1950 live recording, released by Orfeo, but it is not limited to Mahler. But, whereas Scherchen’s excesses could often be infuriating, Klemperer’s are always fascinating.

Comments are welcome