Feature Review: Franz Joseph Haydn – The Symphonies/Dennis Russell Davies

Written by: Antony Hodgson


Symphonies 1-104

Symphony A

Symphony B

Sinfonia concertante, Hob I:105

Stuttgart Chamber Orchestra

Dennis Russell Davies

Recorded 1998-2009 in Mercedes Benz Centre, Stuttgart

SONY CLASSICAL 88697443312 (37 CDs)

A new complete set (issued towards the end of 2009) of the Haydn Symphonies is long overdue. To have completed it over eleven years using public performances as a basis is an unexpected approach but the sound and general balance is surprisingly consistent throughout. The discs are presented in an order influenced by the latest thoughts of some of today’s musicologists as being chronological. This sequence is near enough to the presumed order to make further comment unnecessary. I am also comfortable with the decision to place the Sinfonia concertante at the end of the final disc.

The Symphonies are presented in six groups proceeding chronologically and the accompanying booklet gives sensible titles to each group. In the booklet, the analysis of the music is rather general but is adequate in the circumstances; however, neither the individual movements nor their timings are given. The individual disc-covers are more explicit. These sleeves (about half a dozen CDs for each of these historical periods) are given a different basic colour for each division of music.

Before describing the recordings, it is worth assessing Dennis Russell Davies’s general approach. The most intriguing aspect of this is his use of continuo. With very few exceptions, he uses bassoon as a continuo instrument in those works where no separate bassoon part is written. This is excellent and very much in line with Haydn’s own recommendations regarding performance but the matter of harpsichord continuo brings a problem. Modern performers differ in their approach. Such directors as Trevor Pinnock, Derek Solomons and Roy Goodman use this instrument throughout their performances of the symphonies. Christopher Hogwood (himself a harpsichordist) does not use it at all, with disastrous consequences to the bare harmonies of many of the early symphonies. It is clear that the harpsichord was used copiously at Eszterháza because of the many accounts preserved for the repair and re-quilling of the instruments and we even know that when Haydn went with his band to a neighbouring town he took more than one such instrument for use and for repair still leaving another harpsichord behind for use at court. A document at the time specifically named Haydn as harpsichordist for the touring band. In his admirable Universal Edition of the scores of all the Symphonies, the general editor – H. C. Robbins Landon – recommends use of harpsichord in approximately the first forty or so Symphonies. This is a good rule of thumb.

I wish I could be certain what Dennis Russell Davies required. At first I wondered if he were using harpsichord at all, but then, in the quieter movements, I realised it was there. Then there were louder movements wherein again I thought there was no such instrument but sometimes a single early entry or spread chord by the keyboard player alerted me to its presence before the instrument disappeared into the weight of the orchestra. If I were told that the harpsichord was used everywhere I would not challenge it because it is recorded so extraordinarily faintly that I could never be sure, especially in the fully-scored movements.

We now come to Davies’s approach to instrumentation. In the middle of the 20th-century, H. C. Robbins Landon finally put to rights the boring old fallacy of having the horns in Haydn’s “Festive” C major symphonies put into the lower octave – indeed he was so proud of the results of this particular piece of research that, later-on, he entitled his autobiography Horns in High C. The scores often list the trumpets and horns on the same stave – sometimes stating ‘horns or trumpets’. Often in modern performances trumpets and high horns are both used but a sensible modern convention is to use only high horns with timpani in Symphonies 48 and 82 In the remainder of such symphonies conductors prefer trumpets to double horns only in tutti or fanfare-like passages. Of course timpani must always be used because trumpets (or horns in high C) required the underpinning of drums in the 18th-century – even if this meant that the drum parts had to be improvised.

Dennis Russell Davies takes an extraordinary approach in these works: he never uses horns in C-alto in any of them, choosing always to employ trumpets but also adding the horns at the lower octave where they act as a sort of ball and chain to the brass writing. This invalidates his performances of Symphonies 20, 30, 32, 33, 37, 38, 41, 48, 50, 56, 60, 82 and 90.

Before leaving the vexed matter of the horns, what about the horns used for symphonies in B flat? The answer is simple: these are always in B flat alto. Most conductors agree but Dennis Russell Davies goes further and the horns are ‘Alto’ in symphonies that also include trumpets and drums. This occurs in 98 (for the first time in a recording in my experience) and in 102 (only Leslie Jones and Simon Rattle have previously recorded the work in this way) and the effect is revelatory.

So much for instrumentation – what about Dennis Russell Davies’s approach to symphonic contour? Well the most appealing aspect of his approach to shape and form is his admirable refusal to bow to the worn-out tradition of slowing down at Trio sections of Minuets. Such a movement is normally the shortest in a work and it is ridiculous to have two tempo changes within a short space of time and in any case a minuet is a dance movement and regular pulse is essential. In all the symphonies Dennis Russell Davies permits such a slackening only once – I know it is a special case but it is still not forgivable. I refer to Symphony 95 – the whole Trio is a cello solo – and the sad thing about this performance is that it sounds as if the cellist is ‘slowing down for the hard bit’. Since this is obviously contrary to the conductor’s philosophy, dare I wonder if this solo might have been dubbed in later making it difficult to match the speeds?

Other aspects of approach to contour are consistent – although not always convincing. I really don’t care for the frequent omission of the last two notes of minuets and I also find that the slowing-before and emphasis-upon final chords of movements is done so frequently that it becomes uncomfortably predictable.

A very important feature of a conductor’s approach to the shape of the music is his treatment of repeats. In sonata movements Haydn always marked both halves for repeat until he got to the ‘London’ Symphonies where with one exception he took what would be the 19th-century approach and asked for only the exposition to be repeated. Performers nowadays tend to observe both repeats of sonata movements in the earlier symphonies but later are content to repeat only the exposition sections. Dennis Russell Davies takes a similar view.

When we turn to the minuets, matters become more complicated. In recent times, some musicians have argued that in the 18th-century the internal minuet repeats should be made after as well as before the trio sections. In the 19th-century it became clear that the minuet was played just once through after the trio with some composers (notably Spohr and Mendelssohn) specifying this in their scores. There are good arguments for both theories and Dennis Russell Davies sometimes makes both repeats after the trio and sometimes neither. In those works incorporating a standard Minuet, Dennis Russell Davies makes both post-Trio repeats in about a third of them and observes the conventional ‘no repeats after trio’ in two-thirds. This would be perfectly acceptable but unfortunately in a further eight symphonies the conductor defies all logic by making the first but not the second repeat after the trio. This has a disastrous effect on the proportions and I really cannot accept this ‘three-legged minuet’ structure; for this reason, this means Davies’s performances of Symphonies 21, 39, 42, 68, 63, 73, 78 and 85 unsatisfactory.

Now to the performances themselves.

The Early Symphonies

It is perhaps interesting that not a single one of this early group of symphonies has ever been given a nickname. There is an unfortunate start to the whole project for as early as 18 seconds into the First Symphony I hear what sounds like a clumsy edit, clipping the attack on a chord. This impression is confirmed when the section is repeated because there the same chord is perfectly normal. On the whole I enjoyed the approach to these early works, the light touch in the neglected No.4 is particularly fetching; the high horns in No.5 are played brilliantly and although tempos can sometimes be quite measured, there remains a sense of urgency and freshness. No.20 with its low horns and weak timpani will not do and in No.37 not only are the horns low but the trumpets and timpani are omitted altogether.

The inclusion of the brief Symphony ‘A’ is interesting. H. C. Robbins had the score published by Universal Edition and it was included in Hoboken’s catalogue as Hob.1:107 in B flat. Landon presents plausible evidence for the Symphony – in particular he quotes the Fürnberg Collection in the Budapest National Library, and notes that the MS parts by a professional Viennese copyist has “a few additions in Haydn’s hand” and most scholars seem to have accepted Symphony ‘A’ as genuine Haydn ever since. Two questions remain however: why did Haydn not catalogue the symphony among his compositions and, more importantly, why does Helga Scholz-Michelitsch list the work in her thematic catalogue of instrumental works by Wagenseil? There can be no doubt that this is the work concerned because the incipit that she gives is identical to the first few bars of the so-called Symphony ‘A’.

The First Symphonies written for Prince Esterházy

This is a bigger area, it takes up seven CDs. Twenty-five symphonies are involved and they include a number of works showing a highly original approach to instrumentation. Numbers 6, 7 and 8 – the so-called ‘Morning’, ‘Noon’ and ‘Evening’ trilogy – is hugely popular, the many solo instrumental lines are demanding and altogether they provide a test in style and skill for most orchestras. Their Minuets are notable for the double bass solo in each – played here with strong rhythmic pulse. The strangely shaped No.15 (first movement: slow-fast-slow) is very convincing. In the second-placed Minuet the Trio has delicately played viola and cello solos. No.3, given here as chronologically the 13th, sounds rather earlier but it is a delightfully baroque example of Haydn’s writing and is very well done with excellent horns.

Not all in this group are so consistently good. Symphony 36 is given a decent reading – bassoon continuo does not seem evident – this would not matter too much if only there were a harpsichord to be heard. Anton Heiller was far more dramatic in his première recording of the work around 60 years ago. No. 33 is a write-off (horns in the wrong octave), but the slight No.9 comes decently alive. Symphony ‘B’ (Hob.I:108) is a delightful discovery. Unusually the Minuet is placed second but this is among the best of its period, lightly scored but with driving rhythms and the high horns play superbly. No.14 is excellent too and in No.40 the problem in the Trio as to where woodwind and strings should be used is well-solved because Davies uses the solution given in a lesser-known source in Erzherzog which specifies woodwind only. Stylishness extends to the three-movement No.12 and somehow the hugely slow tempo for the central Adagio makes sense. Numbers 16, 34 and 72 make a good group. 16 is very slight but 34 has hidden depths and I have to say that the omission of the last repeat in the multipart finale makes the shape more logical than indicated in the score – we really don’t need that long section followed by a very final coda to be played twice. No.72 is the partner of the ‘Horn Signal’ (No.31). Both have spectacular parts for four horns. Interestingly Davies uses the timpani part in 72. It may or may not be authentic but when, as here, it’s used sensitively, it is convincing and makes a good contrast with the otherwise-identically scored No.31 which is also well performed with the many difficult solo parts skilfully rendered although in both works soloists are allowed rather wide freedom of tempo.

The performance of 30 ‘Alleluia’ is a great disappointment. It is possible that this is one of the rare examples of the subtitle being given at the time – not perhaps by Haydn but he could well have been aware of it. Unfortunately its brilliant high horns and trumpets, which can be heard in Nikolaus Harnoncourt’s spectacular version, are not employed and Davies does not use the timpani either. A neat performance of No.23 ends strangely. Haydn plays a trick in the finale, ending the second part indeterminately and it is only on the repetition that the listener realises that this hesitant group of notes will represent the end of the piece. Davies seeks to avoid giving away the joke by making the repeat nine bars earlier than marked so the strange non-final notes are only heard once – at the very end. This is a challenging notion and a very good idea but Haydn certainly expected the repeat to be made in full because at the close he wrote a whole bar’s rest: clearly he was gauging the space before the return to repeat at the middle of the movement. 22 and 21 are adequate but both lack rhythmic strength and the character of a pair of cor anglais in 22 – the famous ‘Philosopher’ – is rather tamed in this uneventful rendering.

39 and 28 are not remarkable (and 39 is ruined by a ‘three-legged’ Minuet) but 29 is excellent with much spirit and good detail. Yet again attention is paid to a Trio section and this is a very special example. Haydn writes strangely here: this whole section is a sort of ‘vamp-until-ready’ with dark gypsy-like harmonies but no tune at all. Something needs to be done and Davies solves the problem by having the harpsichord emerge from obscurity to add a sort of basic melody in the top line – very effective. I have heard this done only once before when, in Wilhelm Loibner’s ancient recording, Christa Fuhrmann-Landon improvised in the same way.

The “Storm & Stress” Works (Sturm und Drang)

This important group of symphonies shows Haydn’s genius rising to even greater heights. On the whole the performances do not always do justice to the great intensity of these works and it is sad that the thrilling Symphonies in C with their wonderful high horns and timpani have again to be written off because the horns always lumber along an octave too low. This deprives us of 38, 41 and 48 (Maria Theresia) but there remain some notable moments – In 46 the horns in B are used in the upper octave, very few other conductors choose to do this but I am sure Davies is correct to do so. The whole texture opens out excitingly, and while on the subject the lightly-scored 52 does find the C-alto parts being used properly (the absence of trumpets and drums makes this essential). The remarkable No.65 is another feast for horns (this time in A) and Davies manages to bring out the remarkable sequence of ‘wrong beats’ in the minuet very effectively. The slowish tempos for No.58 also help bring out the strange cross-rhythms – this seems to have been a feature of Haydn’s compositions at this period.

I don’t find the required intensity in the dynamic No.35 and because in the first movement Davies plays the second section only once it means that we are allowed to hear the staggering upward run for horn on just a single occasion – maybe the ‘Society for the Prevention of Cruelty to Hornists’ stepped in. No.43 (‘Mercury’) is given all repeats everywhere – very rare in this work. An excellent version despite a moment or two of imperfect ensemble in the opening movement. No.47 goes well too and here the completeness of the Minuet is particularly convincing because this is the movement which is an exact palindrome, each section is played forwards and then backwards so all those repeats make a perfect mirror image. For those who like employing little-used computer technology, they might like to try playing the whole movement backwards and they will find that it will still come out exactly as written (but you would have to put up with the reverberation coming before the notes rather than after). I was surprised to hear clipped grace-notes in the Minuet of 45, the ‘Farewell’, Davies is usually more thoughtful about such matters, but this is a good dramatic reading and in 44, the ‘Trauer’, the slow ‘funeral’ movement is very touching. This was played at Haydn’s funeral – at his own suggestion. This is a reading from a conductor with a considerable love for Haydn’s music.

Symphonies for Entertainment Purposes

This title might apply to Haydn’s symphonies but I don’t envy the producers having to think up so many titles – this will do as well as any. There are eight discs in this group and once again we have mixed blessings. The same low-horn problem invalidates the grand Symphonies 50 and 56 and it is even worse in 60 (Il Distratto) where there are several fanfare-like moments and the lower octave makes no sense at all. As if this were not enough, a disaster occurs in the Minuet of this work. Here Davies chooses to make both repeats after the Trio – but in this recording, from just before the end of the first part of the returning minuet until just after the start of its repeat, three seconds of music are missing and they are replaced by the same amount of entirely different music from the following slow movement – an alarming oversight.

Symphony 63 ‘La Roxelane’ is a disappointment of a different kind. The original score used high horns, trumpets and drums. This included music culled from other sources, including an operatic overture (for Il mondo della luna) which acts as the first movement – nevertheless it is an excellent and exciting work. In his lifetime Haydn was persuaded to publish a revised version in which he removed trumpets and drums and completely re-wrote the Minuet and finale. This is a much inferior work – the finale is of very poor quality and I wonder if Haydn really wrote it himself. I see no reason why this revision should be performed nowadays but regrettably it is the version used in this Stuttgart set and Davies takes the opportunity to put the horns down an octave in all movements except the Minuet. Symphony 69 is a rare example in this C major brass and drum sequence where a number of experts (but not all) are content with it being played in 19th-century style using low horns. In this case it works quite well like that. It is not one of Haydn’s finest and I have heard it referred to as “a poor man’s Maria Theresia” (there are thematic similarities in the first movement). Davies gives a good solid performance – exceptionally broad yet effective in the opening Vivace and with a notably strong Minuet but the second movement does seem to go on for a long time. The strength of No.66 lies in the superb horn-playing and in 67 Davies’s all-through approach to Trio sections ensures that this weird example with a solo violin over a bass-line that sounds like a hurdy-gurdy is as winning a rendering as I have heard.

The disc containing 61, 53 (‘Imperial’) and 70 provides large-scale music in suitably grand readings Timpani are better in focus in 53 – one of Davies’s best readings and he uses the original finale – many conductors replace it with an overture. Even the harpsichord is given a little more presence. No.61 is just plain strong – this work and No. 53 are scored for horns and timpani but with no trumpets – rare for the key of D major but it works splendidly in both. No.70 is well done – the slow first movement and Minuet nicely counterbalance the weight of the elaborate slow movement which is quaintly labelled ‘Specie d’un canone in contrapunto doppio: Andante’. An outstanding disc.

The remainder of this group is steady and reliable. There is a good, sturdy 75 (I forgive the odd grace-notes in the Trio) and as for the unpopular 62 with its opening movement also serving elsewhere as an overture, this performance has character. Dennis Russell Davies tends usually to be firm and unhurried in Minuets but here he is both fast and effective. Three movements of 73 are excellent but it is a shame that the Minuet is slow and without emphasis and that ruinous ‘three-legged’ disability is imposed again.

Symphonies for the Public at Large

This grouping group takes us right up to the end of the symphonies written prior to the famous visits to London. There are six discs and include the ‘Paris’ Symphonies. By now Davies is accepting the first-repeat-only convention in sonata movements and I confess that in these substantial works this pattern makes perfectly good sense. Only in the finale of 87 do I query this decision because Haydn writes bars containing rests after the final chord. This makes it clear that he expected that second repeat to be made. To consider doubts about the readings first, we can dismiss 82 (Bear) and 90 because yet again horns are an octave lower than required. Something far more bizarre happens in Symphony 83 (La Poule) however. After a calm, easy-going reading of the first three movements, the finale sets out in an attractive way – gracious with gently bouncing rhythm until after three-and-a-half minutes at bar 84 (the fermata on the violins’ held D) madness sets in! Firstly there is a three-second pause until, omitting the up-beat note at the start of bar 85, the music resumes at an incredibly slow pace – ‘Largo molto’ might be an apt description. It then lumbers on lugubriously for a further 19 seconds bringing us to the end of bar 87. Here the sound disappears altogether – not just a cessation of playing but the hall ambience goes completely dead; seven seconds later the music returns where it left off and crawls along for a further twenty seconds until, at bar 93, the original tempo returns and the symphony ends. I cannot imagine what went wrong: maybe the slow tempo sent the engineers to sleep.

The other ‘Paris’ Symphonies have no such eccentricities. It is interesting that the very slow tempo for the first movement Vivace of No.85 (La Reine) seems earthbound yet a similar speed in Kuijken’s excellent recording is convincing. Full marks for Davies’s long grace-notes in the Trio of the Minuet but no marks at all for ruining the symmetry by repeating only the first of the two sections after the trio. No.87 is far more joyful and No.84, which in my experience never seems to find favour with music-lovers, is given a bright, direct reading. No. 86 is good with strong dynamic contrasts in the outer movements and firmly pointed rhythms. No-one ever really hits that splendid fortissimo drum roll at the first-time bar before the repeat in the finale hard enough but Davies is no more disappointing than anyone else.

Numbers 76, 77 and 78 seem to belong in spirit to the previous grouping: 76 is done very well indeed, 77 has a surprisingly fast and rather exciting Minuet (unusually Haydn marks it Allegro) and Davies ensures its effectiveness by, as always, keeping the Trio up to tempo. The dark No.78 remains sombre in character – well-played but not challenging but this is another of those occasions when Davies imposes his asymmetrical repeat system on the Minuet.

What a contrast when approaching the Minuet of No.81 – here Davies gets it entirely right. This Minuet contains one of Haydn’s little jokes when in the Trio he writes out the repeat of the second section in full but puts it into the minor key. Some conductors repeat both second sections all over again – how sensible of Davies not to do so – we don’t need the joke twice over. This is a very good performance, almost (but not quite) on the level of the great one on Mercury with Dorati and the mysterious ‘Festival Chamber Orchestra’ (to be found only on a very rare LP). The highly dramatic No.80 is driven firmly. The main theme of the finale is ostensibly on the wrong beat. Davies stresses the rhythm in an unusually angular way so the ear is even further confused as to where the rhythmic beat lies. I am sure this is intentional and I expect Haydn would have approved. No.79 displays a slightly stolid opening movement perhaps but I like its firmness and here the conductor does avoid his habit of predictable slowing at ends of movements. No.88 is sturdy with admirable strictness of tempo in the tragic slow movement. I had not expected a measured tempo to be appropriate for the finale but it certainly works. There is likeable reading of 89 but it understates the delightful string swoops at the start of the finale.

Symphonies 91 and 92 are interesting. The Allegro assai of the first movement of 91 is extraordinarily slow. There is something about the music that leaves it unharmed – I don’t often want to hear it at this speed but Davies somehow gets away with it – by contrast the finale seems a little breathless – all this can be forgiven for the sake of the Andante – rarely have I heard such superb horn-playing in this movement. No.92 (Oxford) is well interpreted yet under-powered. Timpani is a little reticent, forte chords are not hammered home. Tempos are all acceptable and to the conductor’s credit he avoids the common habit of speeding up for the loud bits in the slow movement, but there are more exciting performances than this.

London Symphonies

The twelve ‘London’ Symphonies are packaged with the Sinfonia Concertante in B flat, the latter a fine showpiece for the Stuttgart soloists – they are all excellent: Benjamin Hudson (violin), György Bognár (cello), Jochen Müller-Brinken (oboe) and Akio Koyma (bassoon). They all combine instrumental precision with rhythmic accuracy. No high horns as with Dorati and Adam Fischer but that is not important.

I am so familiar with the standard numbering of the ‘London’ Symphonies that I listened to them in that order. On the whole these late works are interpreted with conviction. No.93 is admirably solid – a slow tempo for the first movement is always an advantage and a swift one for the Largo cantabile makes sense too. No big exaggeration when at the end of the movement the bassoon confidently finds the note for which all the other instruments had been searching (Haydn’s jokes are so clear that we don’t need the conductor to dig us in the ribs), I like the inflections in the dancing Minuet and my only regret is the slightly understated finale, though the trumpets are exciting. No. 94 is rather good and the unusually fast tempo for its ‘Surprise’ Andante works very well. I also like the broad view of the ‘Allegro molto’ Minuet especially as the Trio flows so well with similarly incisive rhythms. I don’t know why the fierce drum solo in the finale is so backward though – especially as there is nothing wrong with the drums elsewhere in the work. No.95 disappoints because of the soggy Trio section and I could have done with much more drama – I consider this an angry work but Davies takes a smooth approach. Similarly I would have liked far more intensity in No.96 (Miracle). Once again there is justification for slower-than-usual tempos but a lack of punch in forte passages undermines this good intention. We are given the rarely observed second repeat in the first movement.

Symphony 97 is another matter. Textures are clearer and more forward. Much weight is applied and this results in grandeur. If only there were more performances matching this in sheer strength. It does seem from the writing that for once the horns make sense in ‘basso’ because the scoring is very much in 19th-century style. Incidentally the fantastic timpani solo in the Minuet (which is heard four times) is the most exciting I can recall hearing on any recording. The violin solos – written for Salomon (the impresario who brought Haydn to London – are played with panache. This is one of Haydn’s greatest symphonies and in the context of recorded performances in this set; it is one of the most successful.

Number 98 is given a challenging reading. Not quite the ultimate in forcefulness in the big moments perhaps, but those horns in the upper octave are thrilling: Davies convinces me that he is correct in using the instruments in that way. We know that Haydn directed from the keyboard when performing his ‘London’ Symphonies (it was referred to as ‘harpsichord’ but could often have been the fortepiano) and he wrote a famous keyboard solo in No.98. In this particular recording the solo is very distant. The harpsichord is slightly audible at a few moments elsewhere and in the middle of the finale it provides a remarkable set of improvisatory flourishes but it still sounds extremely faint. Just one oddity – whatever happened to the da capo in the finale? This is the only exposition repeat to be omitted in any symphony in the whole set. I cannot believe it was the conductor’s intention – could it be an editing error?

Clearly recorded woodwind is a feature of these recordings and Symphony 99 is the first in which Haydn uses clarinets. Here this change of colouration is immediately obvious. This is a decent, straightforward reading – the observation of both repeats of the Minuet after the Trio gives four opportunities to hear the exciting upward horn run – what a shame that it is understated. On the other hand Davies deserves great credit for doing exactly what Haydn asks in the sudden slowing joke near the end – all too many conductors slow well before the event as if to say “look out, a joke is coming soon!” No.100 is a good test of an orchestra’s mettle and there are moments here when the title ‘Military’ is fully justified – notably in the trumpet fanfares, the following fierce drum crescendo and the bold timpani solo in the finale which is played much more effectively than in the equivalent place in No.94. On the other hand, what feeble high percussion! The triangle player need not have turned up (perhaps he didn’t) and the cymbals never make more than the gentlest splash of sound and sometimes they disappear altogether.

Turning to No.101 (Clock) Davies looks for grandeur but does not come near to reaching it on the scale achieved in 99. There is a contrast of view within the performance with a brisk, crisp ‘Clock’ movement – incisive and exhilarating but it is followed by an immensely slow Minuet – no after-trio repeats this time but the movement still takes nine minutes. Fortunately Symphony 102 soon dispelled my disillusionment and those splendid high horns sound even more exciting than in Rattle’s rather good version. I very much like the long grace-notes in the Trio section, too. Not quite the ideal power in the famous timpani crescendo in the opening movement – but that is par for the course in most recordings.

No.103 brings an element of eccentricity. The opening of this ‘Drum Roll’ Symphony is open to interpretation. Haydn simply asks for a timpani roll. He marks it ‘solo’ and puts a fermata over it – in other words its length is left to the performer. Haydn also adds the word ‘Intrada’. Well clearly this is an introduction of some sort but why this special attention? There is no dynamic mark, so loudness is also a matter of choice. For years conductors had the roll played crescendo-diminuendo but from around the time that Scherchen and Wöldike recorded the piece in the 1950s it has been usual to play a firm loud roll followed by a diminuendo. There is a school of thought however that feels it acceptable to improvise a pattern of timpani strokes. I thought Abbado had approached nearest to madness in this respect, but now that I have heard Dennis Russell Davies I am not so sure! The same effect, with the same instructions in the score also occurs towards the end of the movement. Whereas Davies’s introductory improvisation lasted a quarter of a minute, his second drum roll (or more accurately his spattering of drum strokes) is only half that length but I quite like the way it is led into via a drum crescendo. All this would be fun on a concert-hall occasion but I am not sure I want to hear it every time I return to a recording. I am also worried that the realistic sound of the timpani in their spectacular solos contrasts markedly with the sound of the same instruments when full orchestra enters. They sound as if they have been moved much further back during that short space of time. Nevertheless, this is an unaffected performance and the notion of using the first five horn notes as an introduction to the finale by playing them slowly is a useful way of approaching this unusual passage. For some reason, though initially a little underpowered, the movement becomes more forceful as it progresses and ends strongly.

The recorded performance of Haydn’s last great Symphony perhaps sums up my feelings about the set as a whole – admirable rather than outstanding. This is a reliable and imaginative interpretation and as with many of his symphonic performances the conductor is again able to convince the listener that a slower-than usual tempo for the fast part of the opening movement is a valid way of approaching the music. I find this an appealing element in his music-making. The Andante is brisk and incisively phrased and I was delighted that the beginning of the Trio is played perfectly in tempo. So many conductors stumble over these couple of bars. The finale is unremarkable but suitably faithful to the score.


It is important to consider the three rival complete editions. The reliable Dorati set was especially good at giving suitable textures to the earlier symphonies but the bigger mid-period Symphonies were often bland and understated. Later, things improved and there are some very good performances among the ‘Paris’ and ‘London’ sets.

Adam Fischer’s set was challenging and occasionally controversial. He was prone to decorating repeats in unexpected places and sometimes he reduced the violin line to a solo instrument. He gave many convincing performances but the recording was not satisfactory in the bigger works which could boast excellent string quality but balance was questionable and in particular the timpani was unnaturally remote – indeed at times barely audible. Fischer provided some very decent versions of earlier symphonies in which the sound is far more transparent.

The earliest set to be issued was conducted by Ernst Märzendorfer who displayed great insight. Märzendorfer seemed particularly at home in the more fully-scored works. This is the most interesting and challenging of these rival sets. I regret Märzendorfer’s omission of keyboard continuo, especially in earlier symphonies, but Davies’s instrument is recorded so distantly that his use of it improves only very slightly on its complete absence.


To summarise Dennis Russell Davies’s approach: I approve of his use of harpsichord if that is how he sees it, but I certainly don’t approve of its being so faintly rendered in the recordings. Then there is the sad invalidation of those marvellous symphonies with high horns and timpani (often with trumpets as well) the change of texture caused in lowering the horns by an octave invalidates the purpose of a group of works for which Haydn is particularly famous. The approach to repeats is sensible except for those few works where I have explained that an aberration occurs in the Minuet. Finally there is the overall sound of the orchestra – excellent with notably fine woodwind and brass. Timpani balance varies a bit but at its best these instruments, though modern, are made to sound suitable for the period and usually they avoid being over-warm.

I was very glad to have had the privilege of hearing this whole canon because so many of Dennis Russell Davies’s performances were enlightening. Unfortunately I have one further reservation and it is serious. For some reason applause is to be heard after nearly every work It starts at exactly the same suitable time after the end of each piece and consistently lasts between 30 and 36 seconds, always fading at the much the same place. After a while I found this infuriating. Whether the work was big or small, early or late, the applause sounded much the same every time. This annoyance was underlined for me when two consecutive works on CD 17 (Symphonies 45 and 46) had no applause after them – what a blessed relief! I didn’t manage to catch the engineers using the same applause more than once but then I certainly didn’t listen through all the applause every time. I think this could be a serious dissuasion from purchasing the set, even for those who do not share my reservations about the performances. It seems ridiculous that the time taken up by all the clapping is equivalent to the timing of a whole CD.

This is a bargain set and some offers ask a price for it that equates to around £1.50 per disc. I hesitate to recommend this entire set despite the many excellent things that it contains but the solution for those who wish to seek the best from this project is probably to download the most interesting performances. There are various download offers on the Internet – one that caught my eye was priced at 15 cents per track or 99 cents per album.

Leave a Reply

Your email address will not be published. Required fields are marked *

Skip to content