Aureole etc.

Golden Age singers

Nimbus on-line

Faure songs
Charlotte de Rothschild (soprano);

  Founder: Len Mullenger
Classical Editor: Rob Barnett

Gustav MAHLER (1860-1911)
Symphony No.9 in D Major, Kindertotenlieder*

Janet Baker (Mezzo Soprano)*, Scottish National Orchestra*
London Symphony Orchestra/Jascha Horenstein
(Royal Albert Hall, London 15/9/66 and Usher Hall, Edinburgh 3/3/67*)
Restoration and remastering by Jerry Bruck, Posthorn Recordings
BBC Legends BBCL 4075-2 2 CDs [113.38]

Crotchet   AmazonUK   AmazonUS
 Amazon recommendations

In an interview Jascha Horenstein once called Mahler's Ninth Symphony his war-horse and he certainly rode into battle on it a lot, as Joel Lazar's informative notes for this release explain. In 1966 he gave two performances in London with the LSO. One was at the Royal Festival Hall on April 21st, the other at the Royal Albert Hall for the Proms on September 16th. A recording of the April performance has been available on Music and Arts (CD-235) since 1986 and now we have the opportunity of hearing the September performance in another of the continuing series of Horenstein issues by BBC Legends. This has already brought us Mahler's Das Lied Von Der Erde (BBCL 4042-2), Seventh (BBCL 4051-2) and Eighth Symphonies (BBCL 4001-7) as well as Bruckner's Fifth (BBCL 4033-2) and a coupling of his Eighth and Ninth Symphonies (BBCL 4017-2) and I have reviewed all of them here. His Mahler Seventh is, in spite of inferior sound, one of the greatest recordings of the work with Horenstein's grasping every detail of the music, though it is heard in much better sound on earlier releases by Music and Arts and Descant. The Eighth is in a class of its own and Das Lied Von Der Erde, remarkable for its preparation as well as a penetration to its emotional core few conductors and soloists equal, should not be missed either.

The difference between all of those and this release is that they had been available before unofficially on other labels whereas this particular Mahler Ninth has never before been issued. All that now remain for commercial release among Horenstein's Mahler recordings for the BBC are a Fifth from 1959 with the LSO which can be heard at the sound archive in the Barbican in London and a superb Sixth with the Bournemouth Symphony from 1969 rebroadcast a few years ago. I do hope BBC Legends will issue both of these as well as some Bruckner recordings doing the rounds underground. Horenstein's excellent studio recording of Mahler's Fourth for Classics for Pleasure (574 8822) has just been reissued to go with Unicorn recordings of the First and Third Symphonies as well. This means we ought soon to be able to hear Horenstein in all the Mahler symphonies except the Second. There is a persistent legend that a performance of this with the LSO in South Africa may have survived, but legend may be all it is.

It is still important to stress that this BBC Legends recording of Mahler's Ninth is different from the one on Music and Arts. Not least the fact that it is in stereo whereas the Music and Arts release, in spite of its cover claim, is mono. The 1966 Proms season was the first to contain broadcasts in stereo by the BBC and the source tape for this release appears to have been made off the air on good equipment. In fact so good is the sound that having just listened to many of the broadcasts from this season's Proms I had to remind myself from time to time that this was from thirty-four years ago. Much of the atmosphere you would expect here down to the ripples of applause between movements that I'm pleased to hear re-mastering producer Jerry Bruck has left in. All this contributes to the feeling of the concert hall as theatre that releases like this rely so much on for their success. The value of this aspect above all is my response to people who ask what is the point of such releases over perfectly executed and recorded studio versions meant all along for commercial issue. I maintain that the vivid impression that can come across, as it does here, of musicians in full flow, giving their all to a real audience they can see is irresistible and makes up for passing mistakes that might occur and even be caused by that effort.

We have plenty of recordings that render these works note perfect, many of them worth owning. But there is, for me, nothing to compare with the tensions and risks of live performance before an audience. That is after all how the composer himself would have expected them to be heard. I certainly believe that to compare a recording like this too closely with a studio production is, to some extent, to miss the point of these "live" recordings. In that sense they are different but comparable, especially when they also give an opportunity of hearing a conductor's interpretation of a particular work we wouldn't otherwise have available which is the second important value of releases like this. Far better they are freely and legally available than that they languish in archives, heard by no one, or come to us stolen and bootlegged in even worse sound. No one would be more pleased than I would be if Horenstein had recorded this and the other works in the BBC Legends issues with the benefit of the best sound and the availability of retakes. But he didn't, and Horenstein is a Mahler conductor of such importance, maybe one of the three or four most important conductors ever to interpret Mahler's music, that the fact these recordings exist at all means they cannot be ignored. I would still want to hear them even if Horenstein had, as was once proposed, made a complete Mahler cycle in the studio, as I hope you would also.

From his various recordings it is possible to hear that Horenstein's conception of the Ninth remained basically the same down the years. These two 1966 performances especially show a remarkable consistency in approach even though there are some differences in orchestral execution. In fact there are times when it is only these which enable you to hear this is a different performance from the one we have been used to. The LSO's playing is technically more accurate in the earlier recording than here, but much more of that later. Comparing both these performances with the only studio recording of the work Horenstein ever made for Vox (CDX2 5509) as far back as 1953 I do hear a deeper dimension to the darker side of this work, especially in the first movement. If, as Deryck Cooke maintained, this is Mahler's "dark night of the soul" under Horenstein in 1966 it becomes the "dark dark night of the soul" with no sense that there is anything in this extraordinary landscape other than despair. As a view Horenstein conveys it vividly, without any sentimentality or mannerism, mainly through stressing with savage clarity the lower frequencies in the woodwind and brass. Especially the rasps and snarls Mahler buries in the texture at nodal points that Horenstein digs out like a surgeon cutting through flesh to get to the bone. This is the "dirty end" of Mahler's music that Horenstein's generation knew how to bring out only too well and which many of today's conductors seem unable or unwilling to even countenance. Like Rugby players determined to get to the end of a match without a speck of mud on their shorts. Are they frightened of the true implications? I do find myself concerned that Horenstein nevertheless produces a slanted view of a movement that has some more remitting passages than this performance appears to show. But that is a question of mood, a feeling impossible to pin down with reference to the letter of the score. Horenstein's 1953 account of the movement isn't quite so dark and despairing in this way. This 1966 performance does, however, have the same superb coherence with each section flowing into another with a natural fluency. Not least in the crucial opening pages where the main material is presented in one of the most remarkable passages in all Mahler. A miracle of concision which under Horenstein has the same coherence as the rest of the long movement acting almost as a microcosm.

As so often Horenstein achieves this overarching structural integrity by never promoting any passage above another. A case in point is the passage following the collapse climax at 201-203 that Mahler marks "passionate". Leaving aside the fact that what is passionate and what isn't is a terribly subjective opinion, there is no denying that listened to in isolation this does not sound especially passionate. However, played too "passionately" it might get in the way of the build up to the coming cataclysm that Horenstein is undoubtedly building and which he delivers at 314-318 with overwhelming force, trombones roaring out the fate theme with clattering timps behind. When you get there you realise why he made the decision he did earlier. Too much passion at 201-203 would also detract from the essentially bleak view of the movement Horenstein is projecting. Bertrand Russell said: "Nothing great is achieved without passion, but underneath the passion there should always be that large impersonal survey which sets limits to actions that our passions inspire." One of Horenstein's greatest virtues as a conductor of Mahler was not that he knew how far he could take music like this, but that he knew how far not to take it too. You may disagree, as I do myself, with his conclusions as to the effect this movement should leave you with, but that is the way he saw it at that time, and there is no doubt in my mind that he delivers what he means with stunning clarity. Horenstein was a conductor who could make an orchestra play exactly what he wanted them to and this first movement is a fine example of just that. He faced problems when the orchestra in front of him was not of the first flight, of course, but that is not the case with the LSO who knew him very well.

So all pervasive is the pessimism contained in Horenstein's projection of the first movement that it does cast a pall over the rest and that is a consideration when assessing the whole. I will say straight away that I have heard fewer more impressive performances of the second movement Scherzo than the one here. In the 1930s Horenstein conducted the Ninth in Russia and a young Dimitri Shostakovich journeyed especially to hear it. If the performance of the second movement Horenstein gave then was anything like this you can hear why Shostakovich was so impressed because the Russian's voice can be heard right through it. Or perhaps one should say Horenstein's Mahler is heard right through Shostakovich's Mahler-like passages. Here is Mahlerian irony and sarcasm writ large with the same kind of sound, the "dirty end" palette, that we glimpsed at certain points in the first movement, impressing deeply as it must have done the young Shostakovich. The three tempi Mahler carefully delineates are attended to through a basically steady overall tempo suited to all three, as you would expect in one of Horenstein's distinctive traits, so that as the end approaches it is extraordinary to hear the way he deconstructs the music to its component parts.

The third movement Rondo Burleske follows without a break and this is how it was performed on the night. It certainly seems to me that by doing so Horenstein wants us to notice similarities between the two movements - a stress on ugliness and irony rather than energy and drive, I think. Certainly his similarly steady tempo means that every strand of the music, the counterpoint Mahler is satirising here especially, comes out. However I wonder if the fact that the LSO were not given a chance to retune and rest is partly behind the fact that it is in this third movement that the playing suffers most with slips in ensemble especially noticeable. In music where counterpoint is so important this does present a problem, especially for those who worship at the altar of flawless playing. One glaring fault is in the closing pages where the timpanist enters a couple of bars too soon and stays ahead of the beat for a crucial passage that will make anyone who knows this music sit up. However, to set it in some context it is nowhere near as bad an error as the one in Leonard Bernstein's "live" DG recording (D 201182) between bars 118 and 122 of the fourth movement. There, at the climax of the whole work, where maximum power is needed from everyone, the trombone section of the Berlin Philharmonic simply stops playing. Coming in late or early is one thing, but not coming in at all is a different matter again. I have always wondered where the fault for such sloppy playing should be placed, on conductor or orchestra, but haven't felt it to be my place to lay any. This glaring technical error, like truly "an unsightly carbuncle on the face of an old and dear friend", has never stopped this recording from appearing on many lists of recommendations for this work - and neither should it if Bernstein's more animated view of Mahler is what appeals.

Of course, another view would be that any imprecision in execution only adds to the impression of the "muck and bullets" of a real performance that I mentioned, antidote to the squeaky-clean studio versions. I have sympathy with that, as you would expect. But you do need to be aware that the errors may become wearing on repetition. This, of course, was a year or so before the infamous Royal Albert Hall echo had been tamed by the installation of the huge white dishes in the dome, so it was always hard for the players to hear themselves clearly at that time. I suppose that too might explain why the playing of this movement by the LSO was better five months earlier in the more acoustically dry and echo-free Royal Festival Hall. Irrespective of all that, I actually don't think the movement quite comes off but this has more to do with Horenstein's conception than any vagaries in the orchestra's playing. There is none of the compelling unhinged insanity Bruno Walter unleashes in this movement in another live recording from 1938, for example. The playing of the Vienna Philharmonic there is technically even worse than the LSO's in 1966 and yet the movement is delivered by Walter in a way that I believe must be closest of all to how Mahler intended us to hear it. Like a whole way of life, a whole social fabric is being blown to bits before our eyes. Neither do I think Horenstein solves the perennial problem of how to place the lyrical episode at the centre of the movement that Haitink, for example, manages so superbly in his studio recording or Walter does in his "live" recording with which it is more appropriate to compare this. In the context in which it finds itself here in 1966 the interlude sounds too earthbound, failing to really move or lighten our mood by telling us there is something more noble and consolatory waiting just beyond the horizon. A slower tempo would have helped but the tempo for the main material largely precludes that. It all adds to the feeling I have mentioned before of Horenstein's view of this work as being without hope or consolation. Though, in this case, it may be this was not what he meant. When the main material returns Horenstein also doesn't bring off as effectively the trick of gradually increasing the tempo as he did in the April performance, leading me again to the conclusion that things went better in that performance than it did in this one. What a pity the sound quality we have from that April night is so much inferior.

At this point let me draw your attention to some evidence of Horenstein's meticulous attention to the letter of the score that occurs in his performances of this third movement. At two bars before and nine bars after figure 39 in this movement the clarinets squeal out a return to the main Rondo Burleske material. Most conductors instruct their players to suddenly speed up for these two crucial interjections, sometimes dramatically. Yet there is no instruction to do so in the score, no marking such as "accel." over the notes of the instruments as there is in the piccolos and flutes in the fifth movement of the Second Symphony two bars before 21. In the case of Bernstein's "live" recording on DG the clarinets speed up to such an extent that they scramble the second entry with no score justification to back Bernstein up. Horenstein, however, was almost alone (Mitropoulos the only other exception, I think) in actually playing it as Mahler wrote it. In other words making no, or hardly any, changes to the tempo of the two clarinet entries. However, let us not get too hung up over the minutiae of the letter of the score, please. We are music lovers not adding machines and Mahler himself once remarked "The art of conducting is to play the notes that are not there" which, thank heaven, the greatest conductors have always known. There is indeed the letter of the score, but there is also the spirit of the score. Most important of all there is what could be called the letter of the spirit of the score and it is surely here conductors like Horenstein, Kubelik and Mitropoulos, perhaps the three greatest Mahlerians of all, operate at their best and illuminate the scores in different though comparable ways.

The last movement is almost two minutes shorter in this performance when compared with the one on Music and Arts. However, as he did in 1953 for Vox, Horenstein gives the music plenty of space, never hurrying, always aware of the depth of feeling whilst keeping careful control to prevent the music from spilling over into melodrama and more than counterbalancing the first movement which is not always the case. The final climax leads to a coda that is serene and satisfying. Not quite the numbing experience in the closing pages as in the earlier performance but deeply satisfying all the same. By the way, at the great climax, don't fail to notice the three descending sforzandi on unison violins at bars 124-125 that Horenstein has the orchestra play with just enough separation between them, unlike most other conductors who run them into each other. Another example of a fidelity to what Mahler wrote.

This new release records a great performance of Mahler's Ninth and should not be missed by admirers of this work and this conductor. It's the kind of version needed in any profile of it: hors concours in the pantheon of recordings like Furtwangler's 1942 recording of Beethoven's Ninth. A never-to-be-repeated experience that deserves hearing but which is maybe too truthful to stand many repetitions, inevitably carrying the marks of battle of live performance but wearing them as badges of honour. Like all the best recordings taken from live broadcasts it has the element of tension, of concert hall theatre, that ought to override any passing concerns regarding execution and sound.

For reference versions you must turn first to studio recordings by Haitink (Philips 50 464 714), Barbirolli (EMI 7 63115 2), Walter (Sony SM2K 64452), Klemperer (EMI 5 67036 2) and perhaps Boulez (DG 289 457 581-2). Comparable, though very different as a live recording, is Bruno Walter's first recording in 1938 (Dutton CDBP 9708) which is of equal stature to the ones by Horenstein. But Horenstein in 1966 should be on your shelf and this BBC Legends version of the September performance of that year for choice. If you already have the earlier April performance on Music and Arts my advice is to buy this one too, if nothing else for the fine stereo sound.

The performance of Kindertotenlieder from 1967 is less desirable. For one thing I feel that Janet Baker is not really in sympathy with what appears to be Horenstein's austere view of this work, a view we know from a better-recorded version with Norman Foster on Vox. Baker herself was a much more involved and emotional artist than Horenstein and here she sounds too stiff and uninvolved and the result is a bit flat. I stress what appears to be Horenstein's view of this work because the recording balance of this limited mono tape taken off the air favours Baker over the orchestra so it's hard to really make out what Horenstein is doing to any great extent. There is no distortion to the voice but you will have to adjust your volume control to make sure you feel comfortable with Baker's singing full out. Just eight weeks later she would record this work in the studio with Barbirolli for EMI (5 66981 2). Listening to that straight after this you will hear an object lesson in matching soloist with accompanist. Her view of this work there is the same as Barbirolli's: passionate, empathetic, full of heartbreak and heartache, and I wonder if she was thinking ahead to this that night in Edinburgh. Also the superb EMI sound recording allows us to hear that Barbirolli was certainly alive to every nuance of the wonderful orchestration and I prefer Janet Baker happier with her old friend Sir John every time. I'm glad to have this recording with Horenstein in my collection, though.

Every Mahlerian should own this recording of the Ninth Symphony. It should be treasured by all whose first concern is music and its performance in front of an audience.

Tony Duggan


See also Tony Duggan's comparative review of other recordings of this symphony

The above review attracted attention on

.... And what a wonderfully written and conceived review it is, Tony. This could be
a model for balanced, effective criticism and vivid writing. I came away from
reading it feeling that even if I never heard the performance I could know what
it sounded like from the review. Bravo.
Henry Fogel

WOW! But, yes, I liked the review very much (didn't hear this
Horenstein Ninth yet, only the M & A one).

samir golescu

Brilliant review of an unique recording.
Stephen Hall

Not that I wish to incite controversy (smirk), but it is an interesting exercise to
read D. Hurwitz's review in the light of Tony's review. It's clear (to me at least)
that while Tony may give a bit more of a pass to some poor execution than others
might, David is so focused on execution that he can't, or won't, hear the whole.
And then there's his rather tiresome nastiness, but I suppose we just have to accept
that to get his (sometimes) worthwhile insights.

Bob Harper


Return to Index

Untitled Document

Reviews from previous months
Join the mailing list and receive a hyperlinked weekly update on the discs reviewed. details
We welcome feedback on our reviews. Please use the Bulletin Board
Please paste in the first line of your comments the URL of the review to which you refer.