In a necessarily brief paragraph in the May issue I described Rozhdestvensky's Olympia/Conifer version of this symphony as ''now my first choice''. Very rash: on continued acquaintance the recording quality has begun to wear me down: it is bright, with a close focus on woodwind, harp and xylophone, and with what sounds like artificial reverberation added to the strings at times. But it is a performance of huge power and pungent vehemence and on that account preferable to Haitink's majestic but insufficiently extremist account on Decca. Compare Haitink with Neeme Jarvi's new recording and you will see exactly what I mean by 'insufficiently extremist'; in the hurtlingly fast moto perpetuo for strings which triggers the main climax of the first movement Haitink is a notch or two below the marked speed (in the interest of crisp attack and articulation, no doubt; the passage is indeed superbly played), but Jarvi, pushing his players to the very limit, achieves a quite extraordinary whirling impetus that gives the climax itself a terrifying savagery. After it, the angular string figures seem almost to be twitching with still unreleased energy where with Haitink one is rather more aware of heroically audacious violin-playing.
In the finale, too, Haitink makes less of the alarmingly strange events at the centre of the movement, preferring to see the funeral march and its much later catastrophic outcome as the main argument which those other happenings interrupt. In his hands the catastrophe itself has a Mahlerian nobility, whereas Jarvi, surely rightly, projects it with the utmost bitterness: the alternation of eeriness, uneasy jocularity and temporizing blanched lyricism at the movement's dark heart makes even a Mahlerian tragic outcome impossible.
Bitter' and 'biting' are words that often come to mind while listening to Jarvi (and to Rozhdestvensky), there is a great deal of anger, dissonance and frustration to this music which Haitink and the very splendour of his orchestral sound tend either to understate or inappropriately to ennoble. Ashkenazy does not fall into this trap (his orchestra sound a shade more under pressure than Jarvi's, in fact) and his finale is that much more convincing than Haitink's, but he has real problems with the first movement. His fast overall tempo is difficult to relax from, with the consequence that some passages sound gabbled; others, more seriously, are under-characterized. He is good at evoking the sheer strangeness of this music, less effective at distilling the unquiet emotions that need such strangeness to express themselves, in his performance, in this movement several crucially eloquent passages sound stiffly impassive.
Rozhdestvensky is still unparalleled for sheer venom, bitter vociferation and superbly risktaking orchestral playing; the aggressive brightness of the Olympia recording is tameable, up to a point, but there is nothing that tone controls can do to eradicate the impression that the harpist is sitting on your knee and the xylophone player in your breast pocket. Jarvi gets almost as close to the music's acrid core as Rozhdestvensky does, and his reading is far better recorded. (Kondrashin's fine account for Le Chant du Monde which would otherwise be a contender, has dated sound and is only available as a two-CD set coupled with Shostakovich's first three symphonies.) So Jarvi's would be my first recommendation, and this time I mean it.'