Pioneering Haydn On Record
by Christopher Howell
An occasional series dedicated to commercial and live recordings of Haydn symphonies up to and including those of the first two complete cycles by Ernst Märzendorfer and Antal Dorati
Symphony no. 85 in B flat major Hob. I/85 “La Reine” (1785 or 1786)
A pdf version of this article can be downloaded for offline reading and printing here.
“La Reine” is numerically the fourth of the “Paris Symphonies”. I have in front of me the 1950 Boston Haydn Society edition edited by Jens Peter Larsen, which dates it 1786. Modern scholarship usually places it among the first three of the group, the others being 83 and 87, completed in 1785. The nickname was not Haydn’s, but it is one of the few nicknames that actually appeared during his lifetime, deriving from the fact that the symphony was a particular favourite of Queen Marie Antoinette. It is one of three symphonies in the “Paris” set to begin with a slow introduction. This one opens with solemn dotted rhythms, perhaps a tribute to the old-style “French overture” which inevitably began that way. The following Vivace is a grand affair, exploiting the extra weight of the Paris orchestra, which was far larger than that of the Esterhazy court. A curious Haydn surprise occurs at the point where the orchestra cadences on the dominant, leading us to expect a charming second subject. Instead, the music plunges into F minor with a clear reminiscence of the first movement of the “Farewell” symphony. In truth, there is no “second subject” as such – after this episode, the second subject proves to be a variant of the first. The “Farewell” theme plays a powerful role as the development section opens, followed by a more mysterious working of the symphony’s own principal theme. The recapitulation is truncated so as to omit any reference to the “Farewell” theme. At this stage, Haydn was still indicating a repetition of the second half of the movement as well as the first. The “Farewell” quotation has attracted much comment. If we are looking for a hidden message, this could be the moment in which Haydn bade farewell to his role as a composer with a local base, albeit with a growing reputation beyond the Esterhazy court, and became a “European” composer. But Haydn was not in the habit of using autobiographical ciphers of this kind and neither was any other composer before Berlioz or Schumann. Maybe Haydn knew that the “Farewell” – whether or not so-named – was well known in Paris[1], or even a favourite with the Queen?
The second movement is a set of variations on a charming theme – surely it was this that the Queen found so attractive. It is subtitled “Romanze”, an unusual title for Haydn. As usual with variations based on a theme in two parts, there are numerous repeats – a total of nine – offering a field day to conductors who like to cut such things. As to the origin of the melody, Robbins Landon accepted the assertion by Haydn’s early biographer Carl Ferdinand Pohl (1819-1887) that it derived from a French folk tune called “La gentile et jeune Lisette”. More recently, Daniel Heartz[2] has noted that “Pohl cites no source for the song and none has been found since.” Heartz suggests that the words may have been added by someone to Haydn’s music to create a pseudo folksong, perhaps in order for the Queen to sing it at one of the amateur theatricals in which she liked to take part. The minuet is notable for a chirpy “Scottish snap” motif that is continually interpolated while the trio exploits the effect of the bassoon doubling the violin melody an octave below, over a pizzicato accompaniment. The finale could be described as a monothematic rondo. It is a rondo, that is to say, in the sense that the opening theme returns in the middle and again at the end, but it is monothematic in that the episodes are not new themes but developments of the original one.
Two recordings appeared more or less contemporaneously in or around 1950, and another was made the following year. If the dating 1949/1950 on Remus Platen’s Youtube channel is correct[3], the version by Fritz Rieger and the Munich Philharmonic Orchestra (Mercury MG 10116) would slightly predate that by Erwin Baltzer and the Vienna State Opera Orchestra (Haydn Society HSLP 1008)[4]. In a major survey of the Haydn symphony recordings then available in the USA, the November-December 1952 issue of High Fidelity listed both of them, noting that both were “agreeable in the playing and flawed in engineering”. The Baltzer performance was reviewed by Howard Taubman in the New York Times of 4 June 1950, though in truth his only comment was that it was “welcome”. He noted that the Haydn Society claimed that neither no. 85 nor no. 82 had been recorded before. This does not preclude the possibility that the Rieger was set down earlier, since Mercury, at this early stage in their history, mostly made use of recordings obtained from European radio stations. I have not been able to hear the Baltzer. Information on this conductor is scarce, though sufficient to convince me that he was a real musician not a pseudonym.
Fritz Rieger (1910-1978) was chief conductor of the Munich Philharmonic Orchestra from 1949 to 1966 and of the Melbourne Symphony Orchestra for the 1971-72 season. He acquired a solid reputation as a traditional-style Kapellmeister – on taking over the Munich PO from his predecessor Hans Rosbaud he announced that modern music would be almost entirely eliminated from the orchestra’s programmes[5]. Though mainly active in Germany, he guest conducted in Italy and France. Various recordings from the latter sources as well as Munich circulate among collectors and show he could be, at the very least, an inspiring interpreter of Bruckner. He begins “La Reine” with a strong introduction, after which the body of the movement, if not literally Vivace, unfolds at a good, forthright tempo with sensitive dynamic shading. The first repeat is played. The second movement is taken very gently and tenderly, with all repeats. This is highly attractive and the music does not outstay its welcome, though I suspect that an Allegretto in cut time should really go rather faster. The minuet does seem slow – maybe not so much for the tempo but for the smoothness and the lack of any dance lilt. The finale spins along at a good pace. The High Fidelity writer thought the sound “pervasively metallic”, which is how it sounds on YouTube. There is also some pitch waver in the first movement and a vicious cut-off of the last chord, which may or may not be as per the original LP.
On 14 June 1951, “La Reine” returned to Paris, when the Paris Conservatoire Orchestra set it down under Arthur Goldschmidt (Pathé DTX 105). This can be obtained in a nice transfer from Forgotten Records (FR 2018). Fritz Nathan Jakob Goldschmidt, as he was originally called (1902-1970), was born in Berlin and acquired a good reputation in Germany in the 1920s, including several concerts with the Berlin PO. A series of vicissitudes resulting from his Jewish origins led him to Italy, France and Switzerland. Post-war, he made Paris his base and obtained French nationality. After a strong introduction, he takes Haydn’s Vivace marking at its word, spinning the music along at a brisk one-in-the bar. At first, I was impressed by the general vitality, but it came to seem one-sided, even dismissive, an impression reinforced by the exclusion of both repeats. The music achieved greater stature from the extra space Rieger gave it. The “Farewell” quotation breezes by rather pointlessly here. In the second movement, Goldschmidt is again theoretically right, playing it as a lolloping two-in-the-bar Allegretto, but despatches the music in a matter-of-fact manner – more a gavotte than a Romanze – that does it no favours. The flautist evidently enjoys his moment of glory in the penultimate variation. All repeats are omitted except the very first. The minuet is marginally swifter than Rieger’s, but so humdrum that I almost regretted Rieger. In the trio the bassoon, which might have added character to the proceedings, is scarcely audible. The finale is the best movement, brisk and energetic. This performance is unlikely to reawaken interest in Goldschmidt.
No fewer than three recordings appeared in 1954. I have not been able to hear those by the American conductor Izler Solomon (1910-1987) with the MGM Chamber Orchestra (MGM 63109) or by Karl Ristenpart with the Sarre Chamber Orchestra (9 July 1954, Les Discophiles Français DF 116).
Paul Sacher recorded his version with the Vienna SO on 6 January 1954 (Philips 820184). Sacher (1906-1999) contributed monumentally to 20th century music by channelling part of his immense wealth into the commission of new works from contemporary composers. By and large, if we wish to hear what these works sounded like under his own baton, we have to delve into radio archives. His recordings were mainly limited to baroque and early classical music which, a cynic might say, could go well enough without a conductor anyway. I have heard enough of his work to attest he could genuinely conduct, though he was wont to provide orchestras with a clear guide rather than inspiration. He takes the main part of the first movement at a steady pace where “vivace” applies to the semiquavers, which come out clearly enunciated. There is a more 18th century feel than with Rieger or Goldschmidt. Having adjusted to this, there is some excellent phrasing and the oboe plays his phrases in second subject territory very beautifully. There is a genuine sense of growth to the development and the “Farewell” quotation has proper impact. The first repeat is played. The second movement goes at about Goldschmidt’s pace and has only one more repeat – apart from the first, Sacher repeats the first section of the minor key variation. However, he has a lighter touch than Goldschmidt and phrases with considerable affection, as well as investing the minor key variation with real feeling. His minuet is no faster than the others, but it has some lilt. Perhaps the experience of Viennese players in three-time dance music helped out – they must have been tickled to find the first five notes the same as the famous theme of “The Blue Danube”. With the finale I began, as in the first movement, by finding it rather slow – surely Presto means a little more, but ended up by appreciating the lightness and grace of the performance. I imagine better performances will come up as the survey proceeds, but this is the first I could recommend to anyone.
Two more recordings came in 1957. I regret very much that I have not been able to hear Ernst Märzendorfer’s LP with the Milan Angelicum Orchestra (LPA 5900) given the obvious interest that attaches to this conductor’s work in Haydn prior to his complete cycle.
Joseph Keilberth set down his version with the Bamberg Symphony Orchestra on 11 July 1957 (Telefunken LT 6615, GMA/SMA 20 in the UK). Rather intriguingly, Jeremy Noble, reviewing it for Gramophone[6], began by remarking that “rather to my surprise there has not been an LP version of La Reine in the English catalogues for some time, if ever”. By my reckoning, this was the symphony’s eighth recording, but most were on labels not generally distributed in the UK. It seems a little surprising, if true, that Philips did not try the Sacher on the British market. As it happens, while I enjoyed the Sacher, I find the Keilberth better still. The art of this conductor, once dismissed as a dull Kapellmeister, has come in from the cold in recent years. This has been thanks, above all, to his live Wagner performances and also, perhaps, to his resplendent recording of Bruckner 6 with the Berlin PO. On this showing, his Haydn is not to be sneezed at. His first movement is remarkably similar to Sacher’s in general concept, but achieves greater stature. The relatively light texture of the first forte passage in the Vivace is deceptive, allowing him to invest the “Farewell” theme with fine drama. He is able to play the louder passages with abundant energy and relax in the gentler moments without perceptible changes of tempo. Neither repeat is played, alas. He takes the second movement a mite slower than Sacher, still within a reasonable interpretation of Allegretto in cut time, but with all the gentle, serenading charm you would expect of a Romanze. He is a little more generous with repeats, playing the first in each variation. The minor key variation is touchingly done while the flute is enchanting in the one that follows. There is poetry in the final bars. The minuet goes with a nice lilt and the trio has a Landler-like charm. Typical of Keilberth’s care over detail, the bassoon in the first part is just audible the first time round but allowed to dominate in the repeat. Unfortunately, the second repeat of the trio is missing. The finale is just that little bit faster than Sacher to add vivacity to elegance. As in the first movement, Keilberth finds a tempo that allows him to be tough one moment and relaxed the next. Jeremy Noble thought the playing “honest rather than eloquent” and Keilberth’s interpretation “rarely stodgy and sometimes positively graceful”. It should be evident that I rate it rather higher than that.
Georg-Ludwig Jochum set down the symphony with the Nordwestdeutsche Philharmonie on 18 February 1959 (Electrola SVMP 8017). This is available in another good transfer by Forgotten Records (FR 2011). A degree of interest in the art of Eugen Jochum’s younger brother (1909-1970) has led to the rediscovery of his Bruckner interpretations, which are much more straightforward than those of his romantically inclined sibling. After a solid introduction, G-L Jochum’s Vivace had me reflecting that this movement seems to go slower every time I hear it. G-L, moreover, has very delicate staccatos on the bass line, suggesting a minuet danced on tiptoe. It is time, now, to examine a certain aspect of the phrasing of the first theme. To make this as little technical as possible, the Vivace begins with a four-note falling phrase played by the violins – a long note followed by three shorter ones. In the 1950 Haydn Society edition, the first three notes are marked to be played in a single bow, so completely legato. The fact that there is no staccato dot over the third note means it is not specifically separated from the fourth – the violins simply move their bows in the opposite direction. Rieger and Sacher play the phrase smoothly, as written in the score I have in front of me. Goldschmidt separates the shorter notes smartly. Keilberth has the violins bow just the first two notes legato, separating the other two. However, his staccatos are not extreme and he maintains the overall line. G-L goes further, playing the last two notes of the phrase in a delicate staccato that momentarily halts the music. I found the effect effete and irritating. So where did the staccatos come from? My assumption, given that more than one conductor did more or less the same thing, was that the old Breitkopf score had it that way, together with various other differences of phrasing and notes – but this is the one that changes the character of the music. IMSLP has available the old Breitkopf score, a later Breitkopf reprint, an undated but evidently old edition by August Cranz of Leipzig, Robbins Landon’s Urtext edition and a set of parts issued by J. J. Hummel of Amsterdam around 1787. So far as this phrase is concerned, they are all identical. So the variant phrasing either derives from another edition unknown to me or from a conductor sufficiently authoritative to have influenced a number of colleagues. Regarding the possibility of other editions, to the best of my knowledge the Breitkopf was standard until adoption of the Urtext editions became the norm. I shall refer to this re-phrasing, if it turns up again, as the “Keilberth variant”, since he is the one conductor of considerable stature to have done it this way. I doubt, though, if Keilberth, however well regarded in his day, was deemed such an authority, or conducted this symphony with sufficient frequency, for other conductors to follow his example regardless of the score. Back to G-L. The music loses its way in the quieter passages, especially the development. He invests the forte passages with energy, if not exactly vivacity, but this is not enough to save the performance. No repeats. In the second movement, he adopts a tempo almost as slow as Rieger’s, but with short staccatos that would have worked better at a faster pace. Only the first repeat is given. The minuet is very straight, with precious little dance feeling. The finale is the most successful movement, combining grace and energy at a tempo that, if hardly Presto, is a decent allegro.
1962 saw recordings by a rising star and an elder statesman. Roberto Benzi set the symphony down with the Lamoureux Orchestra on 8 January 1962 (Philips L 02 229 L). Benzi (b. 1937), born in Marseilles of Italian parents, conducted in public at the age of eleven[7] and took lessons from André Cluytens. I have discussed the question of child prodigy conductors in my article on Willy Ferrero[8]. Suffice to say that for every Lorin Maazel or Bruno Maderna, who had adult careers of such significance that their beginnings count as little more than a historical curiosity, there are the Benzis and Pierino Gambas who always laboured under the image of the child prodigy who made good – up to a point. Benzi made a slew of recordings for Philips in the 1960s but his career developed for the most part outside the major centres. The slow introduction to the symphony – which seems to go decently under anyone – impresses by its sense of majesty. He takes the body of the movement at a nice Vivace and manages to express a fair range of mood, perhaps more than Keilberth, from exuberance to mystery in the development. He adopts the “Keilberth variant” for the phrasing of the principal theme. No repeats. Like Keilberth, he takes the Romanze on the slower side, but still within what could reasonably described as Allegretto in cut time. He phrases with a gentle elegance, the minor key variation is expressively done and the flute plays most attractively in the variation that follows. Honours would be even with Keilberth except that the latter at least repeats the first half of each variation while Benzi eliminates the repeats entirely. Benzi takes the minuet more swiftly than any I have heard so far. This is all to the good, even though there is no particular distinction to the playing, but he then adopts a slower tempo for the trio. He follows Keilberth in having the strings dominate over the bassoon initially in the first part, bringing the bassoon to the fore for the repeat. I said above that I did not imagine Keilberth to be such an authority that other conductors would take his recording as a model, but with several points of resemblance here, I cannot help wondering if his record, as the only readily available one by a conductor with an international reputation, was more influential for a while than one might imagine. I do not suggest Benzi is cloning it, but his performance seems at times like a personal reinterpretation of Keilberth’s. He does not omit the second repeat of the trio as Keilberth does, but since he lets the music sag, this is not the advantage as it might have been. The finale seems conductor-proof – Benzi gives a swift, lively account.
Ernest Ansermet made his recording with the Suisse Romande Orchestra on 6 April 1962 (Decca LXT/SXL 6022), as part of the first complete cycle of the “Paris” symphonies under a single conductor. These recordings were appreciated in their day for their unvarnished truth. Reservations mainly centred on the same thing. I find more to appreciate than not here. Ansermet does not inflate the introduction and has the Vivace unfold at a median speed between the extremes heard so far. With clean, rather Mozartian textures, everything falls beautifully into place. This does not preclude either atmosphere or vitality along the way, though Ansermet does not deliberately draw attention to what he is doing. He phrases the opening theme with the “Keilberth variant” making me more curious than ever to know where it came from since Ansermet was hardly one to mimic a conductor twenty-five years younger than himself and the product of a very different school. The first repeat is played. The secret of Ansermet’s art seems to lie in grasping the way in which melody, counterpoint and harmony combine to create musical movement. Lessons learnt, I imagine, from his early experiences as a ballet conductor, from his friendship with Jacques Dalcroze, the founder of Eurhythmics, and from his training as a mathematician. His second movement goes at an easy Allegretto pace that has you moving in time to it. It is not deliberately charming, nicely phrased as it is, but Haydn’s own charm comes across. The minor key variation is sensitively phrased and I only thought that the flute is rather cool and retiring in the variation that comes after this – surprising given that the player is almost certainly the great André Pepin, first flute of the orchestra from 1932 to 1977. Evidently, Ansermet wanted it like this. I think Stravinsky might have enjoyed this account. All repeats are played – and Ansermet quietly shaves almost two minutes from Rieger’s timing with all repeats. Ansermet’s sense of musical movement also just about saves the minuet from dullness, while the finale goes at a good, lively pace.
Ansermet insisted on recording with the Suisse Romande Orchestra almost throughout his career. He was aware that, especially in his later years, it was not technically the greatest band in the world, but he said it gave him the style he wanted. A comparison with a performance he gave with the BBC Symphony Orchestra on 2 February 1964 in the Maida Vale Studio, part of a programme that also included Debussy’s 3 Nocturnes, shows what he meant. There are places where the string playing is cleaner in London, but with a more bluntly drawn bass line the effect is less lithe and transparent. All the same, anybody who attended the concert and who had a copy of Ansermet’s LP but had not primed themselves by listening to it the evening before, might have thought it identical. They would have noticed, perhaps, that there were fewer repeats in the second movement – somewhat illogically, Ansermet omits the second in the first and second (minor key) variations but the first in the flute variation. It might also have struck them that the finale was slower and more cautious than on the record. Another difference is that in London, the flautist gratefully seizes his moment and makes a real solo out of his variation. I wonder which Ansermet preferred? The player was presumably David Butt, who had become principal flute of the BBC SO in 1960 and remained till 1998. This apart, the BBC performance, which only circulates privately, would be welcome enough if Ansermet had not otherwise recorded the symphony, but since he did it is superfulous.
Paul Kuentz’s version with the Paul Kuentz Chamber orchestra, which he founded in 1951, came out in 1963 (Club National du Disque CND 74[9]). This French conductor, born in 1930, achieved sufficient standing to have worked with such artists as Rostropovich and Zabaleta and seems to have enjoyed the benefit of an efficient publicity machine, principally through Disques Pierre Verany. His recording of “La Reine” begins well – phrasing the principal theme of the Vivace with the “Keilberth variant” – and the forte passages of the first movement are taut and lively. In the secondary material following the “Farewell” quotation, he lets the tempo sag, picking up again in the next forte, then sagging even more in the development, and so it goes on. No repeats. The second movement is quite nicely managed at a relaxed Allegretto. He plays the first repeats of the theme and first two variations (up to the minor key one), thereafter none. In the repeat of the first section of the minor key variation, he brings out the viola line as if it were a new theme. An interesting idea, but he does not convince me that Haydn intended anything but a harmonic inner part. The minuet goes at a joyful, lively tempo but, as in the first movement, the moment the music becomes piano, as in the first seven bars of the second part, he lets the tempo sag. He takes the trio considerably slower, beginning with a massively delayed upbeat and picking up tempo during the first bar. He repeats this gambit with further exaggerations at the end of the second part. This might just be acceptable in a live performance, though I would argue that what passes for style in a Viennese operetta becomes clowning when applied to Haydn, but surely no one would want to hear it this way every time? The finale is excellent, taken at a good lively tempo and without eccentricities, but this hardly transforms it into a performance I will return to.
Denis Vaughan’s set of eleven symphonies – the “Paris” plus 88-92 and also the Sinfonia Concertante – was made in 1965 with the “Orchestra of Naples” – actually the Orchestra Alessandro Scarlatti di Napoli della RAI out of contract (RCA Red Seal LSC-6805). I was not terribly impressed by their Symphony 86 but this is a lot better. We have an Urtext edition at last – a good checkpoint for readers with a score is bar 162 in the first movement, where Vaughan plays the A flat in Robbins Landon’s score whereas everybody else so far has played an A natural. No “Keilberth variant”, almost needless to say. The Naples band sounds more like a real chamber orchestra than Kuentz’s, but I would not make an issue of this in the “Paris” of all symphonies since Haydn had a bigger orchestra than this in the French capital. Still, the sound in the Palazzo Reale of Naples is quite full and, in spite of the smaller string section, the engineers have produced a sound as string-based as with any of the full symphony orchestras that have recorded the work. The first movement Vivace is fairly brisk but the effect is lively and trenchant rather than driven. First repeat only. The second movement goes at a very pleasing Allegretto tempo, perhaps more of a serenade than a Romanze but I daresay it would have brought a smile to the Queen’s face. She might even have asked for more since Vaughan plays the first repeat of each section but omits the second. The minuet seems to me on the slow side for an Allegretto, but so have been most others and there is a fair lilt. Given the reduced strings, I would have expected to hear more of the bassoon in the trio. The finale is a lively Allegro rather than the written Presto, but nobody so far has given us a really dashing Presto and Vaughan is as good as most. What is emerging is that this is a remarkably difficult symphony to get right and I am inclined to think Vaughan the best so far.
From a small ensemble in Naples to the mighty New York Philharmonic, with which Leonard Bernstein set down this instalment of his “Paris” Symphonies cycle on 20 May 1966. Oddly enough, in a drier acoustic and with lean, transparent textures from Bernstein, including short, incisive staccatos, the result is more, not less, traditionally classical. Bernstein uses an Urtext edition, so the right note in bar 162 and no “Keilberth” variant as he launches the Vivace of the first movement. He takes a brisk but not overdriven tempo, with some excellent soft playing where needed and a real sense of mystery in the development. First repeat only. The second movement is certainly an Allegretto in two and I began by thinking how charming it is, but after a while I felt the music needed just a little more space to breathe. Original as always, Bernstein takes all the second part repeats in the theme and following variations but none of the first. The proportions seem a little odd. Here as elsewhere, Bernstein seems convinced that minuets should be slow and grand. His rhythmic swing just about keeps it afloat. His treatment of the trio as a Mahlerian Landler is more attractive, though the big delayed upbeat and slowing down the last time round probably needs to be accompanied by his podium choreography to be fully convincing. The finale is one of the swiftest, but with such control of phrasing, nuance and articulation that it does not seem all that fast. So far, nobody has given us the rip-roaring finale that Haydn’s Presto marking seems to imply.
Not long after, in July 1966, Leslie Jones and the Little Orchestra of London were recording their “Paris” cycle (Nonesuch HC-3011). Needless to say, he uses an Urtext edition. He takes the introduction slightly slower than usual, making much of the pulsating quavers from the fourth bar onwards. His tempo enables him to double-dot the dotted rhythms. It sounds fine, though the case for applying this typically French baroque practice to latish Haydn is doubtful – or did Jones feel that Haydn’s Parisian orchestra would still have been double-dotting automatically? Jones takes the Vivace at a quite steady tempo, but brilliant articulation, punchy staccatos and sharp accents in the forte passages combine with strong dynamic contrast to produce a performance of real stature. This is the best conducting I have heard from Jones so far. First repeat only. He takes the second movement a mite slower than Bernstein – just enough to let it breathe more naturally. I wonder, though – and I did also with Bernstein – whether the forte passages are not attacked too strongly, even violently. They are not fortissimo after all and I do not think these are Haydnesque “surprises” – should the sound not be full without destroying the gentle atmosphere? I also thought the flute rather retiring in his/her solo. Since London has such excellent wind players, I can only assume Jones wished it to be played as a delicate tracery, a “bird in the brake”, rather than as a real solo, but it seems to work better the latter way. Still, this is all very nice and every repeat is taken. I realize, as my listening sessions proceed, that what seems to be lacking from every version so far is the sort of vocal quality to the phrasing implied by the title Romanze – even the best performances sound like serenades. We take a certain vocal quality for granted in Mozart’s melodic writing, while Haydn’s tends to be more instrumental in concept. But by using this title, which he did very rarely, was he not bidding to compete in a form in which his friend Mozart excelled? Well, perhaps he failed and the sort of performance I have in mind would not work. Jones is quite slow in the minuet, but with an almost Viennese lift to the upbeats he brings it off well. I still think Haydn’s Allegretto means something a little swifter, but if you have to do it slowly, this is the way. The finale is an outright success – an eager-sounding Presto with good articulation and tight control of dynamics. Altogether, this is splendid – more so than Vaughan’s, I think.
The set of “Paris” Symphonies recorded in 1969 by Günther Wich and the Süddeutsche Kammerphilharmonie of Stuttgart was the fifth “Paris” cycle of the decade. The conductor’s name will inevitably invite the riposte “Günther Who?” but perhaps it was not ever thus. Born in Bamberg in 1928 and currently enjoying deserved retirement, Wich first trained as a flautist before switching to conducting. He held posts in Freiburg (1952-59), Graz (1959-61), Hannover (1961-65) and, most importantly, with the Deutsche Oper am Rhein of Düsseldorf (1965-80 or 81, depending on what reference book you use). He also taught conducting at the Hochschule für Musik of Wurttemberg from 1962 or 1982, again, according to which reference book you use, until 1994. He was particularly noted at Düsseldorf for his attention to contemporary music. By 1966, he was conducting out-of-town concerts in the UK with the New Philharmonia Orchestra, later appearing in the regular seasons of several London orchestras. He conducted Die Zauberflöte at Covent Garden in 1968. In the USA, he conducted Fidelio at San Francisco Opera in 1978 with Gwyneth Jones leading the cast[10]. He also conducted regularly the original-instrument ensemble Cappella Coloniensis, leading them on an extensive tour of the USA and Canada in 1972. Their performance in Toronto of Haydn’s Symphony 86 was judged an “earcatcher”[11].
Association with the Cappella Coloniensis did not necessarily mean a radicalization of the conductor’s outlook, since they had the habit of engaging golden oldies, such as Ferdinand Leitner, or younger conductors such as Gabriele Ferro whose wide repertoire did not particularly embrace HIP. Such practitioners could be expected to draw performances from the group as close as possible to the style and sonorities they were used to obtain from “normal” bands. We should not expect Wich, therefore, with a modern instrument chamber orchestra from Stuttgart, to offer a radical departure from what we have heard so far, nor does he. He might, perhaps, as late as 1969, have used an Urtext edition, but he does not so we have the wrong note in bar 162 of the first movement and the principal theme of this movement is phrased with the “Keilberth variant”. It would be a pity to reject it for this. The introduction has regal majesty and the Vivace, if not especially fast, is superbly taut at a tempo that allows fiery fortes and a spacious unfolding of the softer moments. First repeat only. The second movement is sufficiently songful to have more of the Romanze than a serenade. The flute variation is beautifully played and all repeats are given. The minuet and trio have a gracious flow and the finale, at a good, lively tempo, beams enjoyment. A lovely performance.
My cut-off date generally excludes period instrument performances from my surveys. The coupling of Symphonies 85 and 87 by the Collegium Aureum, however, was set down in 1970[12], though not issued until 1973 in the USA and 1974 in the UK (BASF BAC3006). This conductorless group, led by the violinist Franzjosef Maier, was formed in 1962 and created controversy because, it seemed, it was not controversial enough. The first HIP ensembles tended to revel in raucous sounds and free-for-all textures, the argument being that the poor, unenlightened 18th century players could do no better with their primitive instruments. The Collegium Aureum showed that original instruments or replicas could actually play so beautifully that, cynics said, you scarcely noticed they were original instruments at all, so what was the point of it? As Edward Greenfield remarked, “truth to tell, I confess I should not have noticed the use of such instruments from the sounds alone. If it was the intention to create woodwind tone of the kind Haydn would have heard, then surely the oboe tone is far too smooth and sweet, very much the product of the modern school of playing. Listen to the raucous oboe of the Suisse Romande on the Ansermet version, and there, I fear, unlamented, is the sort of sound that comes closer to strict authenticity”[13]. The EMG Monthly Letter took pride in the fact that it reviewed the records from actual sale copies not review copies, with the result that their somments usually came at least a month after those in other journals. This meant that under cover of anonymity – their reviews were unsigned – they could indulge in a little scholarly bitchery over reviews already published in other magazines. Greenfield was surely the butt here:
The fact that these players have taken the trouble to play old instruments beautifully has caused some critics to complain that their instruments do not sound nasty enough, on the assumption that of course Haydn’s orchestra must have sounded pretty bad, with hideous shrieking oboes particularly. They forget that the nature of the score is such as to indicate that Haydn had in mind many beautiful sonorities such as he could scarcely imagine were his resources so poor. These players succeed in indicating to us what old instruments can really sound like …[14]
Several other features of Greenfield’s review are puzzling. He remarks that “The absence of conductor, reliance on the leader, means that the slow introduction of No. 85 starts very flabbily with slack dotting at a faster Adagio (for safety’s sake) than a conductor would usually select.” Conductorless Haydn performances, albeit with modern instruments, had been issued frequently in the previous decade by the Hungarian and Prague Chamber Orchestras, though not of this particular symphony, so should not have caused surprise. More to the point, the adagio introduction is among the slowest of the thirteen I have heard until now and the dotting is not slack, though it is not double-dotted – nor is there any reason why it should be. Just possibly, Greenfield had a review copy in which the wrong take had been inserted. He also had a “warning for those with perfect pitch: the pitch is authentically low”. On the YouTube version I have heard, the pitch is as near 440 as makes no difference[15].
With half a century passed, I see no reason to judge this performance by criteria different from those applied to other recordings, especially those using small orchestras. Regarding this, the Paris orchestra for which Haydn wrote had 65 to 72 players, including ten double basses. The Collegium Aureum, for this recording, assembled all of 19 players, with a single double bass[16]. Mathematics not being an opinion, one might ask why this is OK for Haydn when 38 players, with two double basses, might be found rather inadequate for Mahler 9, for which about 120 would be considered normal. But I do not wish to make too much of this, since Haydn’s symphonies were widely published and performed in his day and he knew very well that not many centres could call upon a larger ensemble than the twenty-or-so he had in the Esterhazy court.
My reservations concern a slightly different matter. I have noted several times in this series, when a symphony has been played by a conductorless chamber orchestra, that the effect tends to be not so much that of a scaled down orchestra but of an expanded string quartet. A symphony is essentially a public statement. These players sound like a group of friends who have assembled for an agreeable afternoon of post-prandial entertainment among themselves. It is unfailingly pleasant and, above all, unchallenging. Nothing is allowed to disturb what the musicians evidently see as Haydn’s comfortable, untroubled world. But Haydn was doing new things, combining the instruments in new ways, throwing brickbats at old concepts of formal construction. Surely the musicians in the Paris orchestra must have reacted to the originality of what they were being asked to play? This seems to me a clear limit in the first movement. After a sufficiently regal introduction, the Vivace goes at a comfortable pace. Not so comfortable as to exclude a certain decorous energy in the forte passages, but the gentler moments seem too gentle and the tempo eases a little as the movement proceeds. First repeat only. Where their approach does convince, is in the second movement. This has a sweet elegance, even a touch of melancholy that is not confined to the minor key variation. As with Wich, it is more a romance than a serenade and, even more than with Wich, the forte passages are not interruptions. It is as if the singer takes a deep breath and sings louder, but is still singing. It is a pity that only first repeats are taken. The minuet has something of Wich’s gracious quality, but suffers again from their tendency to start well and then let the tempo slacken slightly. This happens again in the finale, which sounds absolutely lovely at the beginning, but ends up a little too comfortable.
“Don’t forget Kurt Sanderling’s Haydn”, wrote a German friend when I mentioned I was preparing these articles. So far, he has not come up because his recorded Haydn is limited to his 1971 set of the “Paris” Symphonies, with the Berlin Symphony Orchestra, and perhaps only two other symphonies. His “La Reine” (Eurodisc 85 965 XK) begins with a rather faster introduction than most, avoiding over-inflation. The Vivace goes at a comfortable tempo though, unlike the conductorless Collegium Aureum, Sanderling keeps it going. I do not know how many members of the Berlin SO were actually engaged but the sound, while that of a full symphony orchestra, is lean and transparent, though string-based. First repeat only. The second movement is not far from that of the Collegium Aureum, maintaining a tone of gentle melancholy and investing the forte passages with a cantabile quality. Strangely, he gives all repeats except the second in the beautifully played flute variation. My heart sank as he began the minuet with a completely legato upbeat. In Sanderling’s hands this does not seem a dance movement at all, rather a baroque arioso in a 19th century arrangement. He does give the bassoon due prominence in the trio, but this only adds to the dolefulness. The finale, if not exactly dashing, has the inner energy I might have expected from Sanderling all through.
So now to the two complete cycles. Ernst Märzendorfer begins with a slow but tough introduction, with strongly pulsating quavers in the lower strings from bar 4. His Vivace is swift and urgent from the word go, with almost jabbing staccatos in the bass line. Things verge on the frantic here and there, but the rightness of this is proved when the music does not steer into the doldrums in second subject territory and in the development, as it tends to even under the firmest of guides at a slower tempo. First repeat only. The second movement is of the serenade rather than the songful type, but well brought off according to its lights. All repeats are played. The minuet is remarkably similar to Jones’s. The tempo is slowish, but lilting Viennese upbeats save the day. The finale is a good, eager Presto. If not quite revelatory, this performance is among the best.
Dorati, like Märzendorfer, begins with a slower-than-usual introduction but, whereas Märzendorfer very deliberately does not double dot, Dorati very deliberately does – fine if you like that sort of thing. As I remarked à propos Jones, this is not baroque music and I am not convinced that double-dotting is applicable to latish Haydn. It maybe that Greenfield’s strictures about the Collegium Aureum’s supposedly too-fast tempo and slack rhythms were made with Dorati in his ears[17]. In truth, Adagio in cut time would seem to imply a faster tempo than any so far tried, and would hardly leave room for double-dotting. Dorati’s Vivace is swift, though more dainty in its staccatos than Märzendorfer’s. The forte passages have splendid vivacity, but Dorati is not quite so good as Märzendorfer at maintaining tension in the softer moments. But this is excellent and some may prefer the greater expressiveness Dorati finds in the pianos. First repeat only. I began by preferring Dorati’s second movement to Märzendorfer’s. The pace is about the same but with a gentler touch that adds at least a little of the vocal quality I hoped for to the serenading. The forte interjections are brusquely despatched, however, and as the movement proceeds, Dorati seems to lose interest. The minor key variation is perfunctory and the flute in the following variation seems uncomfortable with Dorati’s insistence on keeping up the tempo at all costs. All repeats are given. The minuet is alertly shaped after Dorati’s manner, which means with rococo curlicues and bobbing curtseys. Like the double dotting, this is all right if you like that sort of thing. The trio is bland. The finale is neatly turned, just missing Märzendorfer’s eagerness. There is a lot to be said for this performance, and those who like double-dotting in the introduction and elegantly mannered minuets will find still more to say in its favour, but I would not put it quite among the best.
To sum up the state of commercial recordings up to this time, if you want the Urtext and all repeats, the situation is simple – unless one of the recordings I have been unable to hear has the second repeat in the first movement, nobody has played it. For a recording with the Urtext and all the other repeats, the choice seems between Jones and Märzendorfer. As has been seen, several of the textually dubious performances have worthwhile points – I particularly liked the dark horse Wich, and he has every repeat except the second in the first movement.
A number of live or broadcast performances are of interest. The great name of Jascha Horenstein is not much associated with Haydn, so it is curious that we have two performances from him, quite widely spaced in time. The first is with the OSSODRE – Orquestra Sinfónica del Sodre – of Uruguay, from a concert given on 23 April 1955. The later one is with the BBC Northern Symphony Orchestra, dated 11 April 1969. It seems to be a studio production, since there is an announcer but no applause at the end. He uses an inauthentic edition but without the “Keilberth variant”. In three of the movements, Horenstein’s interpretation is remarkably consistent. He begins with an imposing introduction, then takes the body of the movement as a real one-in-the-bar Vivace. So far, only Goldschmidt has risked this, but Horenstein’s far greater control over dynamics and phrasing makes a much better case for this tempo – indeed, he goes a long way to convincing me that this is how it should go. The movement he has radically rethought is the second. Not so much over the repeats – just the first in 1955, just the first two in 1969 – as in tempo. The Uruguay performance jogs along at a serenade tempo, while in Manchester he takes a more relaxed view. The music sounds like a Romanze at last. He finds sad poetry in the minor key variation, not entirely dispelled by the flute’s gentle tweetings in the variation that follows. He takes a broad view of the minuet both times, but with more grace in 1969 and, coming after a much more thoughtful second movement, it sits better in context. Incongruously, the minuet was the longest movement in Uruguay. The first statement of the rondo theme in the finale has two small repeats. I have not mentioned them before since no conductor has seen fit to omit either of them, but Horenstein omits the second in 1955. He takes a fairly steady tempo both times, very pleasant but I would dearly like to know why he took Haydn’s Vivace at its word in the first movement while toning down the Presto of the finale. The BBC performance is available on YouTube and should be heard for the first two movements in particular.
Cameo Classics has issued, as part of its set of Haydn recordings from the Itter Collection, Harry Blech’s performance with the London Mozart Players, given in the Royal Festival Hall on 25 February 1956. The editions used by Blech are something of a mystery. He has here the right note in bar 162 of the first movement, also in the flute part at bar 65 of the trio, another test point. On the other hand, he phrases the principal theme of the first movement with the “Keilberth variant[18]” and replaces many of Haydn’s staccatos in the second movement with legatos and semi-legatos. I am beginning to get the idea that he took the right notes from an Urtext edition and then phrased the music in the way he thought would be most effective. And effective it is, rightly or wrongly. His introduction is impressively regal, and regality continues into the Vivace. It is not as fast as Horenstein’s but it is still brisk and bustling and, above all full-blooded. No loss of momentum in the softer passages, either. Neither repeat is played. The second movement is played as a gentle romance. The legato rephrasings, whatever you think in theory, give it the vocal quality implied by a Romanze. The minor key variation is warmly shaped. Only the first two repeats are played. I have listened patiently to minuet after minuet which goes slower than I think an Allegretto ought to go, and tried to see the point of what the conductor is doing. The only two that went at a good speed, Benzi and Kuentz, took the trio at a slower pace. Blech takes it at a nicely lilting one-in-the bar and maintains his tempo for the trio. And, at least to my ears, how right it sounds. His finale, if not breakneck, is properly eager. Here we have a case of the difference between perceived tempi and real ones. Horenstein’s OSSODRE performance is faster by four seconds, yet does not have this eager quality. All through this series of articles, I have found myself enjoying Blech more than almost everyone else, in spite of his apparent tampering with the texts. My guilty pleasure must, in this case, be curtailed by a few false notes from the horns along the way.
Manuel Rosenthal’s performance with the Orchestre Symphonique de la RTF on 15 December 1958 used to be available as an inexpensive download from INA, but this was discontinued some time ago[19]. Rosenthal uses an old edition with the wrong notes in the first movement and third movement trio, and starts the first movement Vivace with the “Keilberth variant”. His introduction is imposing, while the main body of the movement, a tad slower than Blech, has plenty of energy and he maintains tension in the softer sections. First repeat only. I will mention at this point a question that I should have perhaps followed more attentively all through. This movement, and movements III and IV, are in B flat major so have the horns in B flat. Does this mean they play a single step below the written note or an octave below that? If this were Brahms or Wagner, the answer would be the latter. In Haydn, the general assumption is that they play just a step down. A good many of these recordings have been sufficiently string-dominated for this not to have been a great issue either way, though in more performances than not I did notice the horns raising their voices merrily, if sometimes fallibly, in the second part of the minuet. Rosenthal has them play an octave down. The matter came to the fore with the arpeggio figure that accompanies the recapitulation – they sound like foghorns down there. But to tell the truth, I had noticed a certain muddiness to the textures earlier on and was waiting for that moment to confirm what I suspected. His second movement is rather slow for an Allegretto,yet with chirpy staccatos it has more of the serenade than a romance. The minor key variation is shaped with elegant tristesse and the flautist in the following variation is one of the finest I have heard. He makes the most of the space Rosenthal gives him. Only the first repeat is played. The minuet goes at a good, lilting Allegretto but, alas, Rosenthal takes a slower tempo for the trio. The return to the original tempo for the minuet da capo is particularly jarring. The finale, at a tempo similar to Blech’s, has eagerness and considerable fire in the forte passages. Not quite a top performance of “La Reine”, but an interesting glimpse of a conductor performing repertoire with which he is not usually associated, and with intermittingly impressive results.
Another French conductor, Ernest Bour, is best remembered as Hans Rosbaud’s successor with what was then known as the South-West German Radio Symphony Orchestra of Baden-Baden. Like Rosbaud, he was a strong promoter of contemporary music – his recording of Ligeti’s “Atmospheres” was included in the soundtrack of Kubrick’s “2001: A Space Odyssey” – while also applying an enquiring mind to earlier music. SWR has issued a bunch of Bour’s Haydn recordings, mostly of less-played symphonies. These can be heard via Deezer and the like, as well as YouTube. Unfortunately, it does not seem to have occurred to them that collector-minded people like to have information such as dates and venues. Since Bour’s tenure with the orchestra lasted from 1964 to 1979, I have no proof that all of these fall within my cut-off date, though I have alternative sources for some – not no. 85 – which all date from the 1960s. I have decided to include Bour’s recording in this survey in any case.
I was not very impressed by Bour’s performances of Symphonies 7 and 8, but this is much more interesting. Bour evidently has an Urtext edition, so right notes, no “Keilberth variant” and high horns. I also get the impression that he is using only a small section of the strings, since this sounds more like a chamber orchestra than many that are named as such – allowing the high horns to burst through nicely. He takes the introduction quite swiftly – one of the few to have taken heed of the cut-time marking – and avoids giving it too much weight. Following this, his Vivace seemed surprisingly slow. Bour seems to aim at a Mozartian elegance, but there is vivacity in the forte passages, serenity in second subject territory and mystery in the development. No repeats. The second movement is a little faster than I like, but there is a gentle, relaxed feeling to it nonetheless, the minor key variation has an inner tenderness and the flautist is delightful. Only the first repeat of each section is given. The minuet has a pleasing lilt, maintained in the trio, and there is real eagerness to the finale. This performance is among the best and, if you like transparent, Mozartian textures and do not mind missing quite a few repeats, it may be your favourite.
In their ways, then, Horenstein, Blech and Bour all add a dimension to the commercial recordings.
| I | II | III | IV | |
| Rieger | 07:57 first repeat only | 09:38 all repeats played | 04:54 | 03:07 |
| Goldschmidt | 04:34 neither repeat | 04:25 8 repeats omitted | 04:19 | 03:20 |
| Sacher | 07:43 first repeat only | 04:50 7 repeats omitted | 04:31 | 03:34 |
| Keilberth | 05:44 neither repeat | 06:04 4 repeats omitted | 03:30 2nd repeat of trio omitted | 03:20 |
| Jochum G-L | 06:16 neither repeat | 05:37 8 repeats omitted | 04:34 | 03:25 |
| Benzi | 05:51 neither repeat | 05:24 all repeats omitted | 04:43 | 03:39 |
| Ansermet 1962 | 07:52 first repeat only | 07:52 all repeats played | 04:26 | 03:25 |
| Ansermet 1964 | 08:20 first repeat only | 06:32 3 repeats omitted | 04:33 | 03:34 |
| Kuentz | 05:59 neither repeat | 05:54 6 repeats omitted | 04:17 | 03:19 |
| Vaughan | 08:06 first repeat only | 05:21 4 repeats omitted | 04:40 | 03:30 |
| Bernstein | 08:32 first repeat only | 05:14 5 repeats omitted | 04:36 | 03:13 |
| Jones L | 07:56 first repeat only | 07:46 all repeats played | 04:51 | 03:29 |
| Wich | 08:21 first repeat only | 07:52 all repeats played | 04:38 | 03:46 |
| Collegium Aureum | 08:28 first repeat only | 06:37 4 repeats omitted | 05:01 | 03:25 |
| Sanderling | 08:18 first repeat only | 07:32 1 repeat omitted | 05:08 | 03:20 |
| Märzendorfer | 07:15 first repeat only | 07:33 all repeats played | 04:50 | 03:17 |
| Dorati | 07:27 first repeat only | 06:59 all repeats played | 04:53 | 03:23 |
| Horenstein 1955 | 04:30 neither repeat | 04:22 8 repeats omitted | 05:22 | 03:12 2nd repeat omitted |
| Horenstein 1969 | 06:51 first repeat only | 06:07 7 repeats omitted | 05:37 | 03:45 |
| Blech H | 06:03 neither repeat | 05:39 7 repeats omitted | 03:58 | 03:16 |
| Rosenthal | 07:50 first repeat only | 05:25 8 repeats omitted | 04:37 | 03:20 |
| Bour | 05:54 neither repeat | 05:07 4 repeats omitted | 04:26 | 03:16 |
Christopher Howell © 2026
[1] It had been performed in Paris as recently as 24 April 1784.
[2] Daniel Heartz: Mozart, Haydn and Early Beethoven 1781-1802, Norton 2009, p.361
[3] This recording is not listed in Gray’s Classical Discography
[4] Gray’s Classical Discography dates it 1950 but attributes the performance to Jonathan Sternberg and the Vienna SO, who play no. 82 on the other side. The LP sleeve and disc label can be seen at Discogs https://www.discogs.com/release/20006083-Joseph-Haydn-Symphony-No-82-In-C-LOurs-Symphony-No-85-In-B-Flat-La-Reine
[5] Yet his programmes for RAI included works by Blacher, von Einem and Hartmann so he was evidently open to middle-of-the road contemporaries.
[6] December 1960, p.330.
[7] At the age of six according to the French Wikipedia. Presumably the conductor’s own website is correct: http://www.robertobenzi.com/english/biographie.html
[8] http://www.musicweb-international.com/classrev/2018/Apr/Forgotten_artists_26_Ferrero.pdf
[9] The Wikipedia article on Kuentz claims that this recording was made by EMI but I find no reference to any issue other than the CND and, also in 1962, by Trianon. Kuentz made a further recording of “Le Reine” in 1992, issued by Pierre Verany (PV 730108).
[10] Nevada State Journal 15 November 1978
[11] Toronto Star 6 November 1972.
[12] According to Gray’s Classical Discography.
[13] Gramophone, June 1974 p.51.
[14] EMG Monthly Letter, July 1974, p.2 The author of this review could have been the conductor Harry Newstone, who in later years contributed to the Monthly Letter, or the scholar Basil Lam, whose trenchant views on how not to play the Beethoven symphonies were a longstanding feature of the magazine. Many of EMG’s critics were contracted elsewhere, hence the overall policy of anonymity.
[15] This was “Provided to YouTube by Deutsche Harmonia Mundi”, so is not an amateur transfer of an old LP. It is possible that whoever made the transfer noticed that the pitch was low and “corrected” it. In this case, we must wonder whether the pitch only was “corrected” or whether the speed was adjusted accordingly. Another possibility is that Greenford’s review copy, as well as having the wrong introduction, had been transferred at the wrong speed.
[16] The original LP sleeve, listing the players, can be seen here: https://www.discogs.com/it/release/9571463-Joseph-Haydn-Collegium-Aureum-Zwei-Pariser-Sinfonien-Nr-85-La-Reine-Nr-87-A-Dur/image/SW1hZ2U6MjYyODYyMDM=, viewed 18.3.2026
[17] It is evident from his review that the only recordings of “La Reine” currently available on the UK market were the Ansermet and the Dorati box.
[18] Readers are reminded that I call it this purely for identification. Keilberth’s recording had not been set down at this date.
[19] I am grateful to Massimo Gagliardi for providing me with a copy.
