Pyotr Il'yich TCHAIKOVSKY (1840-1893)
The sleeping beauty, ballet in a Prologue and three Acts, op. 66 (1889) [155:10]
Swan lake, ballet in four Acts, op. 20 (1876) [154:41]
The nutcracker, ballet féerique in two Acts, op. 71 (1892) [84:35]
James Ehnes (violin) (The sleeping beauty and Swan lake)
Robert deMaine (cello) (The sleeping beauty)
Bergen Philharmonic Orchestra/Neeme Järvi
rec. 2012/13 Grieghallen, Bergen, Norway
Reviewed in CD stereo
CHANDOS CHSA5204(5) [5 SACDs: 394:26]
Each of these performances has been reviewed individually on this website over the past few years. While everyone, I hope, would accept that artistic criticism is a highly subjective activity, our reviewers’ reactions to those particular releases - now collected together here as a single convenient boxed set - were nevertheless strikingly polarised.
While, for instance, Dave Billinge clearly loved this recording of Sleeping beauty (“a superb rendering. Neeme Järvi seems never to put a foot wrong"), our mutual colleague Nick Barnard had some pretty significant reservations (review). Similarly, Dave’s enthusiasm (review) for this Swan lake ("another first class ballet release") and, in particular, his suggestion that the conductor "always chooses apt tempi" failed to strike a chord with Dan Morgan who found plenty of instances where maestro Järvi’s speeds weren't to his liking at all (review). And while Paul Corfield Godfrey considered this Nutcracker “a thoroughly enjoyable and very dramatic reading... [and] a magnificently realistic recorded performance” (review), Dan clearly found his store of seasonal goodwill in very short supply indeed (“I wouldn’t want to find this in my Christmas stocking; simply dreadful”) as he once again offered a very different verdict (review).
Of course, it goes without saying that virtually all recorded musical performances may provoke similarly opposed reactions. Indeed, it’s the fact that there is virtually never an entirely “right” or “wrong” response that makes reading and comparing critical appraisals such a fascinating and enjoyable exercise. Nevertheless, criticism of audio-only recordings of classical ballet scores does throw up one or two particular and specific issues that need to be acknowledged – and which have important implications in reviewing these particular performances - before we can proceed further.
We need, for instance, to keep in mind why the music was actually written. Although Tchaikovsky is rightly recognised for his revolutionary introduction of what’s often called a “symphonic” dimension to ballet scores, he nevertheless had no expectation that his dance music would be heard in any context other than the one he originally intended. That, it’s worth emphasising, was as musical accompaniment to professional dancers performing in a staged production - which in turn leads to the self-evident proposition that such music was conceived and was carefully constructed so as to be performed within a number of definite and clearly defined practical constraints.
The primary limitation imposed upon both choreographers - who, in Tchaikovsky’s time, were the ones calling the shots - and composers was, of course, the dancers’ physical abilities. Even the greatest stars might tire over the course of a long evening on stage. A prolonged sequence of material danced at the highest intensity was simply not a viable proposition, therefore, if they were to maintain their stamina - not to mention the state of their feet - to the end of a performance. In order to accommodate that fact, choreographers and their commissioned composers needed to pace their productions, ensuring that for every allegro designated vivo or a con spirito there were plenty of others marked moderato. It’s also worth remembering that, quite apart from the question of maintaining their stamina, not all dancers might have reached the highest performance standards in the first place. Choreographers and composers knew that a ballet would not last long in the repertoire or be taken up by other companies unless it accommodated the capabilities of dancers who weren’t necessarily of the highest technical ability. Once again, therefore, they were wise to eschew too many extreme demands in either choreography or music.
On the other hand, however, we need to recognise that times, society and tastes have changed and that these days many more people will hear ballet scores on CD at home rather than as performed in support of dancers in a theatre. Does that perhaps offer conductors working in the recording studio the justification to transcend those practical constraints mentioned earlier?
The chance survival of an Edison wax cylinder on which Tchaikovsky is heard sharing - in a notably high-pitched voice - a few words with friends and even whistling rather tunelessly for a second or two (Koch Schwann 3-6490-2) proves that he was certainly aware of the early gramophone. Even so, he would surely never have imagined that his full-length ballets might ever take on new extra-theatrical lives of their own by way of developments in sound recording and reproduction. Yet it is without doubt those developments over the past century that have popularised the scores - or at least those numbers included in the familiar suites drawn from them - and encouraged alternative approaches to performing them. Thus, on the one hand, we have accounts from conductors such as Pierre Monteux, Ernest Ansermet, Antal Dorati and Richard Bonynge who have had considerable experience leading orchestras in live performances of classical ballet. Reflective of that fact, their performances still tend to be focused on the requirements of (admittedly imaginary) dancers and are characterised by what we might identify as “theatrical” features - including tempi kept within danceable limits, generally consistent and easily followed rhythms within individual numbers and occasional momentary hesitations to allow for such expressions of individual artistry as a prolonged lift or a dancer’s exaggerated or otherwise emphasised attitude, gesture or flourish. On the other hand, many other recordings have been made by conductors with little or even no experience of accompanying real dancers. In so doing they have frequently paid less attention to ballet composers’ self-imposed constraints and historical performance traditions. By adopting, in particular, tempi that are often so fast or so slow as to be completely impractical for dancers, they are, one supposes, attempting to replace the lost visual colour, glitter and excitement of a staged performance with an enhanced and more variegated aural experience that might, they consider, render the music more immediately appealing and user-friendly to a home audience that may, indeed, be completely unfamiliar with staged ballet and its conventions.
That point leads on to another that's worth consideration - indeed, it may well be the most important of all in the specific context of a MusicWeb review: that contemporary listeners too may approach recordings in one of two very different ways. Listeners who have regularly attended ballet performances will almost certainly be recalling mental images of dancing as they listen to a disc. They may well, therefore, recoil, if only unconsciously, at “unrealistic” speeds such as they’ve never - and would never - encounter in the theatre and which fail to match any recollected stage action that they’re replaying in their heads. On the other hand, those whom we might term ballet virgins will listen to the music with no particular preconceptions and certainly no visual images in their mind’s eye. Consequently, they may well appreciate greater rhythmic flexibility and exaggerated tempi that generate greater excitement or stronger emotional responses to the music in its own right.
I suspect, then, that the differences in critical opinion on the discs under review may well result at least in part from what both Neemi Järvi and his listeners consider the purpose of ballet recordings to be. Should they simply seek to memorialise in sound the score as it might have been heard in a real theatrical performance? Or ought they to explore its further musical possibilities, even if in doing so they risk ignoring the significance of its historical origins? Any answers will, I suspect, depend on whether respondents more habitually watch ballet in live or recorded theatrical performances or listen to ballet music on CD at home. Let me put my own cards on the table and admit that I am personally of the former camp. Nevertheless, I am broadminded enough to also recognise that the medium of sound-only CD does offer conductors an entirely legitimate and even valuable opportunity to escape theatrical constraints and, to an extent, to recreate ballet scores in a distinctive and thought-provoking manner.
As noted earlier, each of these performances has already been extensively reviewed by my colleagues and I refer you to the links provided above for their individual appraisals. Reluctant to complicate matters even more by adding my own detailed critiques to those already made by Dave, Nick, Dan and Paul, I have instead taken the opportunity given by listening to this new omnibus set to take something of a more general overview. I will therefore focus on just a few broad features that strike me as common to Järvi's approach throughout.
From a balletomane’s perspective, the most obvious is an occasional failure to pace the flow of the music as effectively as he might. The stage action - mirrored by Tchaikovsky’s music - has, or ought to have, a dramatic arc, usually beginning low key and gradually building up to a climax which then may or may not be emotionally resolved. Thus, an individual Act often begins with a few moments of scene setting where characters simply move around the stage in a way that accustoms audiences to their roles and to the location and the nature of any imminent action. Regardless of whether it is supporting a staged production or being set down in a recording studio, any performance of the score surely needs to follow a similar pattern of progression, building up to climaxes rather than tackling everything at full intensity from the very beginning. The danger otherwise is that a ballet - especially a narrative one - comes to resemble something of a series of intermittently related episodes, none of which is necessarily of any greater significance or dramatic power than any other. On these recordings Järvi is sometimes inclined to over-dramatise passages where the music is merely supportive of stage action of some sort rather than powering a specific dance. Swan Lake offers a couple of good examples. As the first Act opens (# 1) we find the prince and his courtiers enjoying something of a leisurely picnic. The story makes it plain that the prince feels unfulfilled, both personally and romantically, and that his predominant mood is one of ennui. Järvi surely presses on too strongly at that point, investing the score with a degree of musical energy that simply isn’t supported by the storyline - and, while it’s true that we may not be watching a staged production, we ought still to be conscious of the narrative line that in turn is supposed to be depicted by the music. Similarly, at the opening of the third Act (# 15) when the courtiers are being marshalled for a royal levée and preparations are being made for the arrival of important guests, the extra energy and momentum that Järvi imparts to the music suggests that the evening’s events are already getting somewhat out of hand in a way that should only become apparent later once Rothbart and Odile have arrived. Listening blind to both those passages suggests that something quite dramatic must be happening at the time whereas in fact it isn’t. The same phenomenon may be noticed in this account of Sleeping Beauty. At, for instance, the very opening of the Prologue Tchaikovsky pens a marche (# 1) for a grand assembly of the court at baby Princess Aurora’s christening ceremony. Marked moderato, it is surely meant to convey a dignified ceremony albeit a pleasurable one too. Järvi, however, invests it with a relentless, well nigh frenetic drive that seems quite at odds with the atmosphere that we ought to be expecting. In similar fashion, the festive marche (# 21) that introduces the third Act’s wedding celebrations would, I think, benefit from exhibiting a little more in the way of courtly pomp.
While displaying some consistency in over-egging the scene-setting music, Järvi can otherwise be surprisingly unpredictable in his approach. Sometimes he can be too unrelenting, even rather charmless, as in The Sleeping Beauty # 3/V (the Violente variation) and # 3/VI (La fée des lilas variation), as well as in an account of the well known Act 2 # 17 Panorama that suggests that the prince is making his way to King Florestan’s enchanted palace in a racing car. A similarly driven quality also characterises Swan Lake # 5/IV (the coda of what, since its 1895 transposition to Act 3, has become known as the Black swan pas de deux, heard here in its original setting) or the same ballet's # 16 Danses du corps de ballet et des nains. Neither is The Nutcracker immune to the same approach, as demonstrated in # 6 Scène where what ought to be a moment or two of deeply mysterious magic, the hugely growing Christmas tree, is done at such a pace as to suggest that, in preparation for the family party, the plant in question has been overdosed with a bucketful of quick-acting fertiliser. On a few other occasions, however, Järvi can be almost too expressive. His unpredictable deployment of rubato arguably puts too much weight on individual musical phrases, as can be heard in Swan Lake # 13/II (Danses des cygnes, Solo d'Odette) and the introduction to # 22 (Danse Napolitaine).
In spite, however, of my nit-picking in the preceding paragraphs - and my concerns about overall pacing and dramatic integration - it is hard to deny that the vast majority of the total of 74 individual numbers that Tchaikovsky wrote for these three ballets are performed very well indeed. In particular, Järvi’s accounts of the big set pieces - the Rose adagio, the Waltz of the flowers and so on - are all conceived in a traditional fashion that will please anyone who knows them only through the familiar suites extracted from the full scores. The performances are greatly enhanced by the very skilled playing of the Bergen Philharmonic Orchestra. This is clearly not only a well drilled body but also one which seems to have something of a natural flair for this repertoire. The imported guest soloists James Ehnes and Robert deMaine both make very positive impressions, with Ehnes delivering some particularly affecting moments. Meanwhile, the Bergen orchestra’s own harpist Johannes Wik is deservedly credited on the individual CD cardboard sleeves for his own distinctive - and, in these ballets, not insignificant - contributions.
I must also commend the intention of making these performances as complete as possible. Several numbers in all the ballets are often omitted from stage productions and/or cut from recordings, but may be found in these recordings. That’s the case for some of the aristocratic dances (# 12) that take place during Sleeping Beauty's Act 2 hunt and for Act 3's 17th century pastiche # 29 Sarabande. Even more rarely encountered, but also restored here, is the same ballet’s delightful # 18 Entr'acte, originally written especially for Leopold Auer but dropped even before the premiere performance. It’s also good to hear # 12/VI La mère Gigogne et les polichinelles included - which it sometimes isn't - in The Nutcracker's Act 2 divertissement. The performance of Swan Lake, the score most subject to tinkering ever since its premiere, offers even more substantial restorations in the form of a couple of major numbers originally conceived for specific ballerinas - an eight minutes long # 19a Pas de deux pour Mlle Anna Sobeschhanskaya (more often encountered these days in the form of a stand-alone divertissement choreographed by Balanchine) and a four minutes Danse Russe pour Mlle Pelagia Karpakova. On top of those, we even get to hear a few of the final bars of # 29 Scène finale given a rarely-encountered repeat – something that’s occasionally done in the theatre so as to give the stage crew an extra second or two to mount the final lovers' apotheosis.
My home set-up allows me to listen to these performances in CD stereo, rather than SACD stereo/surround, but in that respect I suspect that I will be in the same boat as most potential purchasers. As we might expect from Chandos, the sound quality of these recordings, engineered by the experienced Ralph Couzens, is very good indeed with all the depth and clarity that these complex, innovative scores require. The layout of the music in CD format has been well accomplished on two of the sets. Acts 1 and 2 of Swan Lake may well reach 81:17 in length but are accommodated on a single disc, allowing the second to hold Acts 3 and 4 quite comfortably. Presentation of The Nutcracker is even more impressively achieved, with the full 84:35 on just one disc, decisively blowing out of the water the old canard about a CDs maximum length being 80:00. Getting a complete Sleeping Beauty on to two discs has proved impossible, however, without a side-break after # 13 Farandole (Scène et danse), about a third of the way through Act 2’s First tableau. Still on the matter of presentation, a special word of praise goes to David Nice, author of the three separate sets of booklet notes. Very thorough and well written, they will be of great interest and use both to those coming new to the music and to old hands.
The Tchaikovsky ballets have, of course, been recorded many times over the years and most readers of this review will, I suspect, already own favourite versions. Asked to suggest individual accounts, I might nominate the Montreal Symphony Orchestra/Charles Dutoit in Swan lake (Decca 436 212-2), the USSR State Academic Symphony Orchestra/Evgeny Svetlanov in The sleeping beauty (Melodiya MEL CD 10 02243) and the Kirov Orchestra/Valery Gergiev in The nutcracker (Philips 462 114-2). Making, however, a more exact comparison with the set under review, if asked to choose a single boxed set containing all three ballets I'd opt for either the National Symphony Orchestra/Richard Bonynge (Decca 460 411-2) or the London Symphony Orchestra/André Previn (EMI Classics 5 73624 2). Both conductors offer colourful and compelling accounts. Moreover, both were recorded in sound that remains first class in spite of having been recorded more than 40 years ago. Pressed to choose between them, I'd go for Bonynge for the extra dose of theatricality that he brings to the scores.
Will these new Järvi accounts prove to be as durable as those others? I really don't know. In the end, however, any evaluation of their worth will, I suspect, depend very much on the individual listener’s outlook. If you are unable to divorce the sounds you are hearing from mental images of dancers performing Petipa’s classic choreography, then these performances will probably not be the ideal choice for you. If, on the other hand, you are one of the many people who love listening to ballet music but are unfamiliar with the dancing - or simply find it not to your taste - you will no doubt appreciate and enjoy the many positive features on offer here from the Bergen musicians and their veteran conductor.
Rob Maynard
