Classics Reconsidered: how does Leonard Bernstein's 1960s recording of Sibelius's Fifth stand up today?

David Gutman and Andrew Mellor spar over the pros and cons of Bernstein’s recording of Sibelius’s Fifth with the New York Philharmonic

The original Gramophone review, from February 1966 ...

Sibelius Symphony No 5

New York Philharmonic / Leonard Bernstein (CBS/Sony Classical)

Given a crack orchestra, it would be difficult for any conductor to fail to make a rousing success of the scherzo and finale of this symphony; these movements are played brilliantly by Bernstein and his New York Philharmonic Orchestra.

I should have thought it difficult not to make a success of the ‘slow’ movement (for it isn’t really slow, since it is marked andante mosso, quasi allegretto), but Bernstein plays it so slowly that it is plain dull. But the real test is the preludial opening movement and in this Bernstein is simply not comparable with Karajan. Karajan’s crescendos, for one thing, are tremendous, while he also seizes every chance of moments of fire and grandeur; Bernstein makes very little of it indeed. As I listened, I thought it pretty boring music and began to wonder how I had ever enjoyed it; but when I put on Karajan the answer was clear.

Bernstein seems much happier with Pohjola’s Daughter, which gets a good performance indeed – but any customer is obviously going to buy the record for the symphony in the first place and in this Karajan is definitely the finer. I was reviewing in the company of a friend who much dislikes Sibelius. He listened to a bit of Bernstein’s record from the start and then returned to his book, only remarking at the end that what he had heard confirmed his dislike of the symphony. I then played Karajan and I noticed that the book was put down – and he remarked that ‘the work isn’t as awful as I thought’!

I have referred to Karajan, but if I were going to buy a record of this symphony, I am not sure I wouldn’t save a good deal of money and go for Gibson, who gives a really excellent and well recorded performance, with the London Symphony Orchestra in fine form. Trevor Harvey (2/66)


The recording reconsidered ...

David Gutman Only Herbert von Karajan and Sir Neville Marriner amassed the kind of copious and varied catalogue associated with Leonard Bernstein in the 1960s. Whatever your take on the iffy critical response Bernstein sometimes triggered over here, his Sibelius did sound particularly harsh on LP. This, the first and most boisterous of his Fifths (there are three in the lists), was made in 1961 and went unissued in the UK until 1966. In the meantime, we’d had Karajan being marmoreal on DG, sonically streets ahead. Then again, Bernstein’s recordings aren’t impersonal, would-be definitive statements. They bristle with life, an aspect curiously absent from Trevor Harvey’s’s listening notes but not, I think, from the music-making.

Andrew Mellor Listening to this brings an actual image of Bernstein conducting to mind. That lies at the heart of what I both like and recoil from in the performance. You have to get used to the idea that what sounds like nature music in most decent performances – the various figurations that freewheel and lock into each other with apparent inevitability – here sounds more like factory music. An extreme example would be the string motifs at around 4'44" in the first movement. They have clearly been shaped and shrink-wrapped in the rehearsal room; there is no way they were thrown up by the musical conversation, organically. Do you hear what I’m getting at?

DG I do, and I can sense that generational divide in expectations. Not having grown up with the cooler more expository Sibelius of Osmo Vänskä and with no knowledge that the piece began life as a sort of modernist extension of the Fourth, we expected a more effortful and interventionist approach, not necessarily a ‘romantic’ one. If you don’t like the way the strings dig in at that point, what do you make of his choices elsewhere in the first movement, not least the raw, blazing coda? You can at least hear the fff timpani entries which others mix down. Perhaps my fondness for Karajan and Bernstein is as irrational as TH’s for Alexander Gibson! I see that the musicologist James Hepokoski has praised their shared ‘“phenomenological” contemplation of Klang’ – whatever that means. They did get slower and more contemplative over the years, though Bernstein had (mostly) yet to embark on that journey at the time of this recording.

AM I think the coda is pure pantomime, but wouldn’t expect less from Bernstein, and in a sense the music can take it.

DG The New York Philharmonic archives include Bernstein’s annotated score in which one can see that he has scrawled ‘Avanti’ where he wants to remind himself to ramp up the tempo (as elsewhere ‘cantando’ when he wants melodic lines to sing). Does that make the coda a stunt?

Bernstein conducting Sibelius's Fifth with the LSO in 1966, the same year that the New York PO recording was reviewed in Gramophone

AM For me, the problem isn’t lack of reserve or overheating, it’s simply the technical handling of the motifs. TH doesn’t like the slow speed in the central movement – but why? I would suggest it’s because, again, Bernstein isn’t standing back and letting the river flow; he’s choreographing – and choreographing interlocking motifs that need to find their own equilibrium naturally, musically. This is the lesson any Sibelius conductor or player needs to learn, and it hasn’t been here. Isn’t this one difference between Mahler and Sibelius – that Mahler needs to be machined or powered, whereas Sibelius needs to be blown or carried?

DG Well, I don’t see the point of writing a theme and variations in which continuity is prized over contrast. I know it’s always been fashionable to think of Sibelius in terms of organic growth, but he was also a master of movement, colour and even charm (belying the stern image partly invented by his wife and her nationalist chums). I’ll admit there’s nothing poco about the poco largamente towards the end of the movement; and that outrageous final ritardando was imposed by all the old-school conductors, Karajan apart. Otherwise, I can’t see that Bernstein lacks authorisation for what he does. He brings out allusions to Viennese classicism in the delicate hustle and bustle and makes the Brahmsian woodwind writing more Brahmsian as well as ‘indulging’ the romantic element. The playing is anything but dull. My guess is that the dynamic range would have been severely compressed on TH’s original LP.

AM You’ve avoided my larger point about the style!

DG Bernstein was the first Western conductor to record all seven Sibelius symphonies in stereo (the Japanese having got there first) and I’m not sure how well the New York Philharmonic will have known the less familiar works, if at all. But the point you make applies to all his music-making to some degree. According to producer John McClure, ‘With Bernstein, we would listen to playbacks and then we’d definitely have to make changes … He was always trying to highlight stuff to make sure no one missed the point of this solo or that solo – sometimes to the detriment of the overall perspective.’

AM I agree entirely; this seems to be Bernstein’s modus operandi, and your quote underlines why his performances often sound fissile – he clarifies the moving parts so they bounce off each other. Great for Mahler and Nielsen. Dangerous (not in a good way) in Sibelius. But I’ll concede that where it does work wonderfully is in the final pages, as the theme fights and twists its way through, against, the contrary voices and harmonies. But how do you – as a critic who, I think, appreciates the long structural game – react? Can we really revel in this final struggle if it hasn’t been adequately prepared?

DG Gramophone’s long-serving Sibelius expert Robert Layton regarded this Fifth as ‘one of the finest of its day’ (3/90), and even TH notices that it’s ‘played brilliantly’. That said, I know many people hear Bernstein’s work your way. Can we at least agree that at this stage of his career there’s no case for confusing those enthusiastic, didactic impulses with self-regard? The ‘competing’ lower layers of Sibelian texture were given greater prominence in his DG remake, perhaps to point up how the composer’s hard-won processes create, in Robert Simpson’s terminology, ‘a vast slow motion of their own, like that of the sky as the earth rotates, while upon the planet’s surface there is teeming human and animal movement’. But, oh dear, it’s almost always painfully slow. Give me the sheer physical excitement as generated in 1961 – the finale at least represents Bernstein at his best.

AM What a beautiful quote from Simpson. He makes the point that you allude to, which is that the tempo in any Sibelius symphony from the Third onwards is arguably a single one, a single idea; you don’t stitch the sections to each other, you create a gravitational speed from which all the adopted tempos stem. Agreed, Bernstein isn’t being self-regarding; I love his honesty and his gregariousness with this music. But he hasn’t mastered that concept, whereas others – before him, even – did. So either way, it is his personality.

DG Bernstein may not possess the mastery of transition that allowed Karajan to give the impression of a gradual and seamless increase in tempo all the way through the second (scherzo) phase of the first movement, but if you really want to time travel, don’t forget Robert Kajanus and the LSO in 1932. Sibelius’s drinking partner would seem to exemplify a spontaneous performing style more flexible, erratic and human than Bernstein would ever have countenanced. The first movement ends with a similarly precipitate, but here chaotic, sprint. Authenticity, like ‘classic’ status, is not fixed.

AM Ha ha, yes – you can take the boy out of Manhattan, but … Of course authenticity isn’t fixed or ring-fenced, but it’s funny how conductors from the north, from Kajanus to Berglund to Okko Kamu to Vänskä to John Storgårds, make their own freneticism feel more natural and somehow slower.

DG Well, here’s one Vänskä fan who won’t be dumping his Bernstein!

This article originally appeared in the January 2018 issue of Gramophone. To find out more about subscribing to Gramophone, please visit:

Gramophone Print

  • Print Edition

From £67/year


The Gramophone Digital Club

  • Digital Edition
  • Digital Archive
  • Reviews Database
  • Events & Offers

From £90/year


Gramophone Reviews

  • Reviews Database

From £67/year


Gramophone Digital Edition

  • Digital Edition
  • Digital Archive

From £67/year


If you are a library, university or other organisation that would be interested in an institutional subscription to Gramophone please click here for further information.