Toscanini was probably the first conductor to record the Beethoven symphonies to the accompaniment of a publicly stated agenda; the old, unstated premise ''Beethoven: a personal view'' replaced by an evangelist's determination to give us the truth, the whole truth and nothing but the truth. It was not an historicizing agenda, though it was an anti-romantic one and, to some extent, anti-German. Nor could it be easily replicated; though in 1962 Karajan and a rejuvenated Berlin Philharmonic came close to mirroring it in what was the gramophone's first integrally packaged subscription set of the Nine.
It was the arrival of period performance that reintroduced the idea of a public agenda. And what an agenda! At times, the list of items to be considered—of wrongs to be righted—has looked more like a supermarket check-out roll than Toscanini's simple sheet of Milanese vellum.
The new Archiv set conducted by John Eliot Gardiner—a remarkable set, that I suspect many will rate as Mr Knightley rates Emma Woodhouse ''faultless in spite of her faults''—comes with its own lengthy agenda. The tone is set by Peter Czorny's booklet essay ''In the Spirit of Rediscovery''. The recordings are offered, he tells us, in the hope of transporting the listener back ''to that moment when this music burst forth into a world of heroes, wars and revolution, creating its own world of the sublime and ineffable''. It is a theme that is developed by Gardiner himself in a characteristically robust and contentious 20-minute talk on the project that comes gratis on a sixth CD.
At one point in his talk, Gardiner suggests that Beethoven wanted his musicians to live dangerously. To push them to their limits, he avers, ''was part of his whole aesthetic purpose''. It is an idea that has some peculiar consequences in the new cycle. A live performance of the Fifth Symphony—a performance that defies the moderating influence of many of Beethoven's written tempo markings and even occasionally upgrades what are already fairly urgent metronome markings—burns more or less consistently at white heat. It is a tour de force. Elsewhere, though, the approach generates a terrible feeling of musical enforcement. Yet when Gardiner stands back from the fray—feeling the music urgently but naturally along a pulse that is his own and the orchestra's, as he does in the Seventh Symphony—the results can be finer still. The new Seventh is not merely a tour de force: it is, by any reckoning, a great performance.
Sometimes—as in the First and Fourth Symphonies—the compulsion to live dangerously sits cheek-by-jowl with wiser counsel. The metronome mark for the slow movement of the Fourth, quaver=84, is palpably too fast and Gardiner ignores it in favour of the relaxed but by no means staticquaver=69 favoured by conductors like Karajan (in 1962), Toscanini and Weingartner. The result is playing of great tenderness and imagination. The passage cited by Gardiner in his interview on page 14 (Adagio, bars 60-68, 5'17'' ff) is especially beguiling.
Roger Norrington, whose approach to these scores seems to me to reveal a keener feel for the everyday dress and furnishings of period Beethoven, treats the Adagio as an exquisite period dance. He is quite close to the printed metronome; but on grounds of style, rather than dogma.
It is very different, though, in the case of the Andante cantabile con moto of the First Symphony. This movement (metronome mark quaver=120) has generally been played at too lugubrious a tempo (Furtwangler 84, Toscanini a leisurely 96), though Karajan broke the mould in his 1962 recording with a beautifully flowing quaver=104. Norrington, as urbane here as in the first movement of the symphony, manages mm 114; Harnoncourt, nailing his colours to the fence, settles for mm 108, a charming but by no means inexpressive reading markedly similar to Karajan's. By contrast, Gardiner is uncompromising. He goes for mm 120. The result is a route-march; a delectable serenade is turned into marionette music.
The modern preoccupation with Beethoven's famously fallible metronome marks at the expense of his famously exact written tempo instructions has long baffled me. Even where a half-decent contemporary performance was a possibility (an assumption that rules out most of the Ninth Symphony), Beethoven was now so deaf he could check neither the accuracy with which his calculations had been made and conveyed, nor their effectiveness in performance.
The finale of the Fourth Symphony is a fairly representative battleground. The marking is Allegro ma non troppo; quick but not too quick. (Minim=63-6 might be a rough translation.) Beethoven's printed metronome is minim=80, an Allegro molto. A simple miscalculation or a cruel piece of musical revisionism? It is impossible to say. I don't know whether Toscanini was the first conductor to opt for out-and-out speed here. He was certainly the most influential, but even he only brought it off successfully on rare occasions. Thus, though his 1939 BBC SO recording (Biddulph, 5/94) is a miracle of elegant high-speed articulation, his 1951 NBC version sounds desperately hard-driven. Karajan's 1962 recording is another miracle of high-speed articulation, beautifully poised; but he never tried to repeat the trick; the 1977 remake (DG, 4/88) is a buoyant but non troppo minim=66.
The new Gardiner is very quick indeed, exciting but joyless. Of course, the fact that he can articulate the music at all at this speed is not only a tribute to his players; it also partly validates his claim that a 60-strong ensemble of old instruments is capable of achieving exceptional clarity of rhythmic and instrumental detailing. It is purest nonsense, though, to claim, as he does, that these attributes are unique to old instruments. Toscanini's 1939 Fourth is just one of a long list of recordings one could cite that prove otherwise. ''What you lose in opulence, you gain in transparency and rhetoric,'' says Gardiner. I don't know whether the Vienna Philharmonic under Karl Bohm (DG, 12/91) can be branded 'opulent', though it has to be said they paint the Pastoral Symphony in richer colours and in greater detail than Gardiner does here. Oddly, in a set that is for the most part stunningly well recorded in a variety of shrewdly chosen locations, the first movement of the Pastoral has rather a lot of murkily distant wind playing. (Several crucial key-changing entries in the development virtually go by the board.) And what of those many seemingly otiose details of orchestral colour with which Beethoven so carefully peppers his score? The ten bars of pizzicato crotchets he gives to the second violins at bar 376 in the recapitulation of the first movement, for example. It is Bohm not Gardiner who, like the old countryman in Thomas Hardy's Afterwards, notices such things.
Of course, it is naughty to seize on such a detail; yet the sometimes hubristic claims made by Gardiner on behalf of his own music-making do occasionally call for modest qualification. In general, his orchestra, which shares some musicians with Norrington's London Classical Players, is a virtuoso band capable of astonishing feats of sonority and articulation. The fact that one or two of these performances are live is an earnest of how much things have progressed in the last decade. Which is not to say that work of the LCP under Norrington sounds dated. There are moments when the LCP's solo wind playing has a bloom and finish to it that isn't entirely matched on the Gardiner set. Nor am I any more convinced here than I was in Gardiner's recording of Brahms's Deutsches Requiem (Philips, 4/91) that the string sound we have here in certain espressivo or cantabile passages is either agreeable or authentic. The awed sotto voce in the Pastoral Symphony's hymn-like coda sounds very prim. Nor do the first and second violins have quite the weight of tone one ideally needs in the flashing swordplay of the Seventh Symphony's final bars. There is nothing 'opulent' about the playing of the Vienna Philharmonic under Rafael Kubelik or Carlos Kleiber (two conductors who divide their fiddles antiphonally—something that, astonishingly, Harnoncourt declines to do). The Vienna sound is, in fact, extraordinarily spare; but it also tougher, more resilient, than that of the ORR.
Inevitably, the new set makes several claims on the textual front. (Jonathan del Mar is Gardiner's principal adviser.) These, too, may occasionally seem peremptory to collectors who have a wider knowledge of extant recordings than the present team either possesses or requires. Not that this in any way invalidates the sharpness of Gardiner's insights or the keenness with which they are executed.
Two examples are worth citing, both from the Ninth Symphony. In his talk Gardiner illustrates the marvellous diminuendo effect—the sound retreating into the distance—of the timpani interventions near the start of the second half of the Scherzo. It seems that what Beethoven wrote here were not accents but diminuendo marks. Gardiner calls this a revelation. It is certainly not often heard, though anyone who has Kurt Masur's 20-year-old Leipzig reading of the Ninth will be familiar with the effect. (Masur once did a good deal of first-hand scholarly research on Beethoven—see his 1968 essay ''Autograph und Druck'' in Musik und Gesellschaft.)
Gardiner has also taken a bold initiative on the question of the tempo for the tenor's Alla marcia in the finale of the Ninth. Quite simply, he takes the metronome 84 to be a dotted minim, not a dotted crotchet. Again, this isn't a new idea, nor does Gardiner actually take 84 as his figure. To judge from recorded performances, the 1937 live London recording (EMI, 9/85—nla) in particular, Furtwangler—one of the work's finest interpreters—took a broadly similar view of the problem. That said, I haven't heard the Alla marcia and the subsequent effortless graduation to the fugato realized with such logic and panache as it is here. Norrington took the opposite solution, conducting the Alla marcia at crotchet=84 as a painfully slow march, with a dirge-like fugato. Hardly an Allegro assai vivace. But, then, I prefer Norrington in the Trio of the second movement. Where Gardiner gabbles, Norrington, like Klemperer (EMI, 8/90) before him, treats the Trio as a pastoral interlude, a view that the orchestration seems to validate.
Symphony No. 1. The opening is superbly judged. Gardiner doesn't overplay the Adagio molto, thus avoiding the over-romantic, world-weary feel of Harnoncourt at this point. The Allegro con brio, always something of an awkward customer and often played with a fatal languor by members of the old German School, is pretty quick. Like Toscanini, Gardiner regards the metronome mark crotchet=112 as achievable. In practice, there is a slightly conscripted feel to the playing, especially in the second subject group where the phrasing sounds pinched. At a slightly slower tempo, Norrington brings out the unenforced joy of the music, its kinship with the dance. After his absurdly brisk reading of the second movement (see above), Gardiner goes on to conduct dazzlingly successful accounts of the Scherzo and finale.
Symphony No. 2. This was the first symphony to be recorded and is very fine throughout, as indeed are the rival performances by Norrington (the coupling of Symphonies Nos. 2 and 8 was a Gramophone Award winner) and Harnoncourt. By following the written tempo markings and his own musical instincts Gardiner produces a perfomance of the first movement that, if anything, opens out the drama even more compellingly: the slow introduction measured, the allegro con brio brimming with energy, self-evidently elated by its own new-found reach. In the Larghetto, he doesn't attempt what is apparently a faulty metronome mark (quaver=92) nor does he render the movement interminable in the German style. Toscanini's flowing songful account (mm=80) was always an agreeable solution; but given finely nurtured playing mm=84 can produce an ideal blend of motion and repose as Karajan demonstrated in his 1977 cycle (DG, 12/87). This is Norrington's tempo, and Gardiner's. Harnoncourt is slower in this movement, and more mannered.
Symphony No. 3. More revolutionnaire than romantique. A very fast first movement gets within spitting distance of another impossible metronome mark. That and keen texturing undoubtedly make for a tremendous sense of dramatic urgency. There is also an unerringly placed climax at bar 671 (15'04''), a crucial point of arrival, ignored by many interpreters. Unfortunately, there is also too little accommodation en route of the rich cargo of ideas that Beethoven has shipped into this movement. Gardiner's rough-and-ready preparation for the recapitulation is a case in point. The imaginatively charged bridge passage for solo horn and flute over a pizzicato bass needs to be played with the utmost poise. Toscanini has it, not least because his basic pulse (mm=52) is brisk but never hectic. In their haste to get to the recapitulation itself, Gardiner and his players are decidedly unpoised. Gardiner says he abandoned an attempt to play the Funeral March at the printed metronome mark. Still, he isn't far off it, and it remains pretty blank and inexpressive. Norrington is equally quick in these two opening movements, but he has greater litheness in the first and makes more of the strange mood of the Funeral March in the second. Both conductors are superb in the last two movements; but these are considerably less than half the story where the Eroica is concerned.
Symphony No. 4. An unusually quick introductory and brisk Allegro vivace. Gardiner treats the pivotal drum entry before the recapitulation atmospherically—closer to Karajan than to Norrington with his Gothic, dynamically exaggerated death-rattle. Glorious slow movement, impossibly quick finale.
Symphony No. 5. Here is the stuff of which revolutions are made. Where Harnoncourt takes a rather old-fashioned view of the piece and where Norrington tends to overplay his hand with yet more unscheduled interventions by gung-ho timpani and brass, Gardiner plays the piece pretty straight, and at white heat. The orchestra is superb, helped by the Francophone bias of its sound base. That said, the Scherzo (which has its repeat) is surely too fast. It starts briskly and not especially quietly (Berlioz said of the opening should ''fascinate like the gaze of a mesmerizer''). At the entry of the main theme at bar 19 the horns blaze away at a tempo faster than the metronome or the Allegro marking. The pace drops back for the Trio, which is just as well since the strings are hard-pressed to articulate clearly. The finale is also very fast, again ahead of what is generally regarded as a good metronome. The only other great conductor I recall doing this is Szell. I am not sure this is the whole story as far as the Fifth is concerned. There is a grandeur to the Scherzo-cum-finale, over which Beethoven laboured so long, that could be seen to reflect a vision (Hegelian, to be precise) that transcends the politics of revolution. Still, for its eclat terrible, this is unbeatable. The slow movement is also superbly shaped and directed.
Symphony No. 6. Despite some lovely playing in the slow movement and a general air of brisk efficiency, this is a rather joyless account of the Pastoral. Nor is it at all a spiritualy uplifting one, as Klemperer's is (EMI, 8/90) or Giulini's (Sony, 5/94). The scherzo—''A merry gathering of country folk''—is a very high speed affair. At such a pace the various amusing false entries rather lose their point; to play in this village band you would need to be a virtuoso, and teetotal to boot. I am reminded here of an old New York PO player who said that when Mahler conducted this movement it was jolly, whereas Toscanini was just waiting for the storm to come up.
Symphony No. 7. A glorious performance. The introduction sets the scene with an ideal blend of weight and anticipation. The Vivace has a splendid dance feel and a power that is utterly unforced. Scherzo and finale are also superbly paced. The Allegretto is eloquent with a sense of barely sublimated grieving. Marvellous brooding basses and fine, veiled colourings. The recording is magnificent, finer than the Norrington which has a rather foggy finale, The Archiv recording was made in All Hallows, Gospel Oak: empty, there is a splendidly open 'ring' to the acoustic.
Symphony No. 8. In the Trio of the third movement, the slightly tentative wind choir sounds as though it has been left in the vestry of All Hallows. In general, the symphony thrives on the Gardiner approach, though in the finale the emphasis is again on high-speed locomotion. Gardiner even outpaces Karajan (1977—DG, 8/87), the previous holder of the course record. In neither case is the sound of the respective string sections 100 per cent focused and 'seated'. (Compare the superior focus and articulation of the 1962 Karajan version.) 72 bars a minute, Harnoncourt's tempo and Karajan's in 1962, is quite fast enough here. In this movement, metronome chasing merely fazes the players and foreshortens the listener's perspective on the movement's huge architectural reach.
Symphony No. 9. I have never heard the first movement dispatched as rapidly as it is here, not even by Toscanini. This is another example of a dubious metronome (crotchet=88) being preferred to a very specific tempo marking: Allegro, ma non troppo, un poco maestoso. In fact, Gardiner doesn't get the bit between the teeth until bar 51, so the celebrated introduction has room to breathe. (Norrington is the reverse; he begins at a disquieting crotchet=88 but then proceeds to modify the pace during the course of what is a far more lyrically and dramatically various reading of the first movement than Gardiner's.) Of course, Gardiner isn't entirely inflexible and he and his players show remarkable skill in making busy detail 'tell'. Yet a lot does go by the board. For instance, the exquisite and wholly unexpected interlude in the coda (bars 469-77; 11'01'' - 11'14'') when horn and oboe play a kind of pastoral plaint. Vaughan Williams saw here an anticipation of the finale's joy theme; ''Suddenly the clouds lift and a mirage, like a vision of joy, appears for a brief moment''. Not in Gardiner's performance, where the placing of horn and oboe is vague and ill-focused. The great French conductor Francois-Antoine Habeneck (much mentioned by Gardiner) who pioneered accurate and responsible Beethoven interpretation in the 1830s is said to have brought a singing quality to the intricate thematic detail of this astonishing movement, a movement that orchestras in places like London and Dresden had hitherto gabbled mercilessly.
The slow movement is also played very quickly. Perhaps there is a point to this, but I fail to see what it is. As both Klemperer and Toscanini have demonstrated, it is possible to play the movement swiftly and unsentimentally, yet feelingly too. However, Gardiner's finale is superb. Tempos are unerringly chosen, the choral singing is beyond criticism, and there is a rare expressive quality to the singing of the solo quartet. Still, superb as Gardiner's account of the finale is, neither he (driven first movement and near-meaningless slow movement) nor Norrington (eccentric finale) can be said to conduct a wholly satisfactory Ninth. Harnoncourt, by contrast, conducts a performance that is—slightly untypically—more or less without eccentricity, consistently fine.
Given the quality of the playing of the Orchestre Revolutionnaire et Romantique, the often exceptional Archiv sound and the single-mindedness of Gardiner's approach, I wasn't surprised to find at the end of my three-week stint of listening that his set emerged ahead on points over the Norrington and Harnoncourt. Which is not to say that he would get my vote as the 'period' choice over Norrington, whose EMI set, for all its several oddities, has something of the evocative power of a Balzac novel. (Balzac once said Beethoven was ''the only man who ever made me feel jealous''.)
And if that is an accolade Gardiner might expect to covet: well, it strikes me that he is at his finest—in the Seventh Symphony or the Second—when he is out of costume, competing, no holds barred, with the big boys. At best, the physical and intellectual vitality of his music-making brings us close to the Ding an sich, the inexplicably wonderful thing-in-itself. It is a best that occurs only intermittently in the new set. That it occurs at all is perhaps a sufficient miracle.'