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. 103 in E flat major Hob.I/103 “Drumroll” (1795)

A pdf version of this article can be downloaded for offline reading and printing here.

The first problem with this symphony is – what to do with the opening drumroll? Haydn marked it “solo”, in one sense an obvious comment since a single timpanist is required anyway. Perhaps the idea was that the player had a certain liberty to do what he liked with it. The autograph has no dynamic marking. A footnote in the Robbins Landon edition shows that a contemporary copy (1797) marked it fortissimo, followed by a diminuendo, while another had a crescendo and a diminuendo but no other marking. What seems to be the first published edition (Pleyel c.1802, available at IMSLP) has the crescendo-diminuendo without a dynamic marking. The long standard Breitkopf edition (c.1854, also available at IMSLP) has the crescendo-diminuendo but has added, presumably on the editor’s own initiative, a pianissimo marking. Thus for years, this opening was usually made to emerge from silence, increase very slightly then disappear into nothing. Some authentically-minded conductors have rejected this romantic interpretation and begin with a healthy fortissimo crash. This apart, the old editions differ in smaller details of phrasing and dynamics which do not affect the music too much, though it is always better to have these things cleaned up.

A few considerations might be made about the second movement. Firstly, the tempo: Andante più tosto allegretto, literally “A somewhat lively pace”. Roger Fiske, in a review I will quote more fully below, tells us that Haydn originally wrote just Andante, then added the rest for fear it would be taken too slowly. Nevertheless, the old Bärenreiter edition had only Andante so this is what the older generation of conductors grew up with – a generation, moreover, that had a Brahmsian tendency to treat andantes and adagios as practically the same thing[1]. Added to this, there was a tendency to play andantes marked 2/4 as four-in-a-bar not two. Theorizing apart, many of the performances that treated this movement as something close to a funeral march demonstrated that it has enough substance to take it, while not all those that trip along charmingly, regardless of the minor key, avoid diminishing it.

The other question is, do you play all the repeats in this movement? There are seven. This is a typical Haydnesque double variation movement, alternating a theme in the minor with one in the major. Both themes are in themselves simple binary statements with each half repeated. That accounts for four repeats. Next comes a variation on the minor key theme, and this has the first half repeat written out, since Haydn varies the orchestration for the repeat. If the first part was not repeated in the first statement of the theme – a few conductors omit even this – it seems rather lopsided to have the section repeated in the variation. The second part is not repeated in this variation. Then follows a variation of the major key theme, led by the solo violin. Repeats are indicated for both parts. If the attractiveness of this variation, or the ego of the first violin, induces the players to observe at least one of these repeats, it will again be lopsided if the repeats were not observed on the first statement, but this has not worried some conductors. That now makes six repeats. Next comes a dramatic variation on the minor key theme, with a repeat indicated – this is the seventh repeat – for the first part but not the second. After this, there is a further major key variation leading to an assertive conclusion, but no more repeats. At the old slow tempo, this movement takes around twelve minutes with all repeats. Only Golschmann and Dixon offer this, but both prove that it need not be too long. Swifter allegretto-oriented performance have whittled it down to nine minutes or less.

The minuet has no marking except “Menuet”. The fullness of the orchestration led earlier conductors to play it in a grandly ceremonious manner, and the best of them also managed an underlying minuet lilt that kept it afloat. Nobody has tried to turn it into a scherzo, but swifter one-in-a-bar performances have become more common. Of the finale, I only mention that, before publication, Haydn removed a few bars just before the end. If played, these considerably undermine the triumph of the final coda. Märzendorfer plays this ending. I understand that, on the original LPs, Dorati had a supplement that included the finale with Haydn’s first thoughts, but I have not heard this.   

If you do a Google search for the first recording of this very popular work, AI intelligence will tell you that two notable early recordings were the 1958 Jochum and the 1963 Karajan. Such are the times we live in. My dim-witted natural intelligence was able to pool the entries of WERM[2], CHARM[3] and, especially, Mike Gray’s Classical Discography[4] to produce the following list – I hope fairly complete – of the performances issued up to and including the cycles by Märzendorfer and Dorati.

Golschmann Saint Louis SO 18.3.35
Ludwig Berlin PO 19.3.41
Heward Hallé O 24.9.41
Solti LPO 29+31.8.49
Heger Bamberg SO issued 1950
Scherchen Vienna State Opera O 10.50
Munch Boston SO 26.12.50
Beecham Royal PO 29.1.51
(Seegers Berlin SO 1952)
H. Blech London Mozart Players 17.1.55
Remoortel Stuttgart Pro Musica O 1.6.56
Wöldike Vienna State Opera O 1.6.56
Jochum Bavarian Radio RO 20.2.58
Beecham Royal PO 9.5.58
Dorati Philharmonia Hungarica 6.6.1958
Wand Cologne Gürzenich O 1959
Markevitch Lamoureux O 23.12.59
Paumgartner Salzburg Mozarteum 1960
Stein Dresden Staatskapelle 16.3.60
Karajan Vienna PO 9.4.63
Ptáčník (artistic director) Prague Chamber O 9-10.67
Jones L/Little Orchestra of London 1968
Maazel Berlin Radio SO 14-18.10.68
Bernstein NYPO 10.2.70
Märzendorfer Vienna Radio O issued 1972
Dorati Philharmonia Hungarica 27.11.72

The performance by (possibly) Heinrich Seegers is in brackets because it was a Royale LP (1329). This company dealt largely in pseudonymous recordings, mostly pirated from existing issues, many of which have been identified, others taken from radio tapes, unidentifiable if the radio station no longer holds the original. No conductor of that name is known to have existed.

A performance said to be by Hans Swarowsky with the Vienna State Opera Orchestra was issued by Tuxedo in 1992. I can find no reference to an earlier issue of this recording. Swarowsky has been the subject of numerous misattributed recordings, though in truth they usually give themselves away by naming a non-existent orchestra. John Hunt’s Swarowsky discography does not include this, so I prefer to be cautious, though to be fair, he does list the Symphony no. 100 with which it is coupled. .

In addition to the commercial recordings, the following live performances seem of interest:

Jochum Hamburg State Philharmonic O 17.1.43 (issued by Melodiya, later Dante Lys)
Abendroth Leipzig Radio SP 29.10.51 (issued on Tahra)
Schmidt-Isserstedt Hamburg Radio SO 1954
H. Blech London Mozart Players 3.9.56 (issued by Cameo Classics)
Newstone Haydn orchestra 8.8.59 (issued by Cameo Classics)
Rosenthal French National Radio O 22.3.60
Matacic Orchestra Alessandro Scarlatti della RAI 1.10.61
Martinon Chicago SO 12-13.1.67
Dixon Hessian RSO 25.10.1968

It seems strange that such a popular symphony had to wait until 1935 for a recording, by Vladimir Golschmann and the Saint Louis Symphony Orchestra, but I find no reference to an earlier one. A soft drumroll, with barely a crescendo-diminuendo leads to an introduction that is mobile while not lacking in ominous dark colour. The Allegro con spirito has infectious verve and the repeat. Some of the articulation is sizzling. Remarkably for a 78 recording, the second movement has all repeats, and a leisurely tempo. There is considerable distinction to the phrasing and the pawky grace holds the attention to the end. The minuet is grand, but the playing has a lift that not all similarly slow versions manage. The finale goes with great spirit – the repeated notes sound eager rather than hammered out and there is no suggestion of tub-thumping. This first recording did Haydn proud[5]. 

I have not been able to hear the 1941 Leopold Ludwig version. The first to be made in the UK was by Leslie Heward and the Hallé Orchestra. Heward (1897-1943) was a conductor of whom much was expected, but his early death leaves us with little to judge him on. His pioneer recording of the Moeran symphony has remained famous. Apart from various concerto accompaniments, this Haydn seems to be the only other work of significant proportions we have. In the first movement, he allows the timpanist more space to make a crescendo-diminuendo than Golschmann did, and the ensuing introduction has an air of mystery, even of morning mists. We note immediately the sweet timbre of the Hallé winds. The really soft, light playing of the strings as they take up the Allegro con spirito surely declares a conductor of uncommon gifts and the whole movement, at a marginally slower tempo than Golschmann’s, has a wonderfully joyous, open-air feeling. We must also note the excellent condition of the Hallé strings at this time. The repeat is taken. In the next movement, five of the seven repeats are omitted. This choice may have been forced upon the conductor, enabling the music to fit on two 78 sides against Golschmann’s three. His gentler, warmer conception has a degree of wistful poetry Golschmann did not achieve, thanks again to the fine wind players. The minuet skips along one-in-a-bar. I had remembered finding it too fast, but this time round I found it so delightful that I prefer to banish any suspicion that this was done to make it go on a single 78 side. It sounds joyful, not pressured. The finale has a few romantic tricks. The opening horn motif is played below tempo and there are rallentandos here and there. The repeated notes are so light and airy as to suggest proto-Mendelssohn fairy music while the forte passages are exuberant yet refined in texture. On this showing, Heward was one of those rare conductors who could somehow release the music from the orchestra rather than force it out of them. What a pity he left so little.

Very little crescendo-diminuendo in Georg Solti’s initial drumroll – he allows a little more when it returns later (Decca AX 333-5). The wind in the introduction do not have the sweetness of Heward’s and the chording is slightly imprecise – the 1949 LPO appears to have been a less refined body than the 1941 Hallé. The introduction proceeds with more care than character – Solti seems mainly interested in creating a single long line and in this he succeeds. There is no particular magic as the Allegro con spirito enters but it is crisp and bright, just a little bullish in the forte passages. The repeat is played. Solti manages more repeats than Heward in the second movement – just three down – but his faster, though not hurried, tempo still fits onto two 78 sides. He treats the movement as a slow march, quite nicely sprung if you like it that way, but rather one-sided. In the solo violin passages, the player’s natural instincts attempt to expand with a little freedom here and there, only to be brought firmly to heel by the conduct’s rigid beat. The minuet is a grand, ceremonial affair, impressive in its way, but Solti does not characterize the trio, simply playing it politely. The second repeat of the trio is omitted, evidently in order to fit the movement onto a single side. The finale goes with keen energy, becoming a touch rabid in the forte passages.

Robert Heger’s version with the Bamberg Symphony Orchestra was issued in the USA – and seemingly nowhere else – by Mercury (MG 10084, 1950). The label states that it was recorded by Bavarian Radio. Heger keeps the crescendo-diminuendo to a minimum then proceeds with a moderately slow introduction that impresses by its sense of latent power. The Allegro con spirito is given fairly steadily but there is a feeling of jubilation in the forte passages. Interestingly, the quavers move at about the same speed as the crotchets in a traditional performance of the scherzo from Beethoven’s “Eroica”. The repeat is played. The “Eroica” comparison continues in the next movement, where Haydn’s C minor has all the gravity of a funeral march. The C major theme suggests a cortège, maybe depicting the dead hero’s hearse being carried home for burial. There are Beethoven parallels, too, in the wailing wind during the next C minor section. I wondered at the beginning if the music would bear such gravity, but it not only does so, it emerges as a remarkably powerful statement. Just two repeats are omitted – the second in the violin solo variation and that in the following C minor section. In this context, a ceremonial minuet has sense and Heger makes more than many of the trio, offering a moment of quiet contemplation. Jubilation is again the key to the finale. The repeated notes look ahead, here, not to Beethoven 5 but to the finale of the “Eroica”. Heger (1886-1978) was usually seen as a Kapellmeister, an upholder of tradition rather than an original interpreter. Was this symphony traditionally played in this way in Germany?

In this light, Hermann Scherchen’s Westminster recording (WL 50) may be not so much sui generis as a personal gloss on the type of interpretation exemplified by Heger. The last two movements are remarkably similar, with a certain stiffness to Scherchen’s minuet, compensated by a fiery – rather than jubilant – finale. Scherchen begins the drumroll with a strong accent – though not a fortissimo crash – followed by a diminuendo into silence. The introduction is incredibly slow. From my experience of Scherchen, I was prepared to discover that he could make even this work but I have to conclude that he goes beyond the limit of what will hold together. Presumably his equating of one crotchet of the introduction to one bar of the ensuing Allegro con spirito is intentional – a case of the theorist getting the better of the artist. The Allegro itself is similar to Heger’s though with less generosity to the phrasing – the jubilation sounds forced. He omits the repeat, which creates an interesting situation. Given his very slow tempo for the introduction, the movement contains almost the same amount of slow music as fast – the introduction plus its reprise before the coda amount to 4 and three quarter minutes against five and a quarter for the Allegro. I daresay this was intentional too. Did he see the movement as essentially one of ominous brooding which the lively music is unable to dispel, leading inevitably to the funeral march of the next movement? If so, I can only say that, for me at least, he fails to make his case. The second movement is treated, as with Heger, as a funeral march, but with shorter staccatos it sounds bleaker. It is also somewhat slower, or seems so. There is an iciness to it that holds the attention less well than Heger’s heartfelt grief. I came to find it rather tedious. Only two of the seven repeats are played – those of the first statement of the C minor theme.

After a gentle drumroll, Charles Munch (RCA Vic.1200) offers a fairly mobile introduction, seeking out the more graceful turns of phrase and keeping the shadows to a minimum. The Allegro con spirito has plenty of energy at a brisk but not pressurised tempo. Though the Boston orchestra sounds large in its famously reverberant hall, Munch’s textures are sufficiently transparent for it not to sound like Beethoven. The second subject is elegantly played with the unusual feature that Munch treats the appoggiaturas as short acciaccaturas. The repeat is observed – not a given with Munch and in fact the next movement has just one repeat – the first. At Munch’s delicately strutting tempo, Haydn’s C minor sounds quizzical rather than tragic and some of the wind interjections later on sound positively cheeky. This is a logically correct interpretation of Haydn’s Andante più tosto allegretto marking. The minuet is slowish but with an elegant lilt that removes the ceremonial element, and the wind phrases that close the first part of the trio sound quite impertinent. The trio is very elegantly done. The finale has plenty of ebullience and fire, but without heaviness. A likable performance.  

I complained that Sir Thomas Beecham’s rather rococo version 1936 of Symphony 93 reflected more of his own genius than Haydn’s, though his later versions moved in a different direction. Here, too, in his 1951 “Drumroll” (American Columbia ML 4453), his agenda seems to be to banish for ever the myth of “Papa Haydn”. His opening drumroll makes only a small crescendo-diminuendo, but manages to sound extremely threatening. The introduction is slow, but there is tension from the start and, on the way, Beecham cossets lovingly any major cadences he can find. His Allegro con spirito is fairly steady, but with tense, terse articulation, not over-indulging the second subject. No repeat. His second movement is firmly in the funeral march category, if not as slow as some, with wailing wind and dramatic fortes. Only the first repeat is taken. The minuet is slowish but has a galumphing air that removes the ceremonial aspect. The trio is expressively phrased, avoiding mere elegance. Second repeats are omitted in both the minuet and the trio. The finale is not especially fast, but it is urgent and fiery, carrying all before it. In spite of the shortage of repeats, this is the first performance so far that has me reaching for the word “great”. And yet … there was a humanity to Heger’s performance that I do not find here.

The 1955 recording by Harry Blech and the London Mozart Players (CLP1066) appears to be the first by a scaled-down orchestra specializing in this repertoire, though the combination of “traditional” string playing – no brittle staccatos – and a fairly reverberant recording produces results that are far from small scale. The opening drumroll begins, if not fortissimo, with a strong accent followed by a long diminuendo. The introduction is mobile and builds up strongly. The Allegro con spirito is driven fast. Like Munch, Blech treats the appoggiaturas in the second subject as acciaccaturas. At this swift tempo, the result is perky, rather like a cheeky little boy whistling the tune in the street. There is an infectious rhythm to this performance. The repeat is played. The second movement looks back to Heward in treating it as a Schubertian ramble, now sad, now pleasingly pastoral, with no suggestions of a funeral march. The wind interjections as the music returns to the minor key express gentle regret. The two repeats are played in the first statement of the minor key theme, all others are omitted. The minuet goes at a lilting one-in-the-bar tempo only marginally slower than Heward’s. The finale goes at a swift tempo with plenty of spirit but not quite Beecham’s rising tension. Interestingly, I was convinced that Blech was faster than Beecham, whose performance I described as “not especially fast”, but the stopwatch tells me he is a few seconds slower. This may be partly due to Beecham’s drier recording, but Beecham had a control of articulation and phrasing that only a very few conductors could attain. Nevertheless, Blech produces plenty of excitement without tub thumping.

Recorded live a year or so later – included in the Cameo Classics/Richard Itter collection – Blech’s interpretation, choice of repeats included, is virtually identical. However, with an audience in attendance, there is an extra touch of urgent peremptoriness and, above all, a greater sense of overarching line to the performance. Furthermore, whatever the critics said about the Harry Blech show as it rolled on into the 1970s and beyond, his orchestra was in fine form back then and he sounds at one with them. Apart from a couple of dodgy horn notes near the beginning of the introduction there are no lapses in the live performance that might jar over time. This is no-holds-barred Haydn, but it did not have me relating it to other later composers (Beethoven or Schubert), or even to Mozart – just to Haydn. 

There is a tenuous connection between the Belgian conductor Edouard van Remoortel (1926-1977) and the conductor of the first “Drumroll” recording, since van Remoortel succeeded Golschmann as conductor of the St. Louis SO in 1958. It was not a happy relationship. After Golschmann’s amiable but effective methods, van Remoortel was a neurotic perfectionist who tried to dismiss 42 of the players and, probably worse still, insisted in holding post-concert rehearsals to correct things that had not been “all right on the night”. By 1962 he was gone. The “Stuttgart Pro Musica Orchestra” is a typical Vox orchestra (PL 9860), probably a pickup band of mainly moonlighting players from Hans Müller-Kray’s Stuttgart Radio SO. There sounds to be a smaller complement of strings than that of Blech’s London Mozart players. The initial drumroll is pianissimo with scarcely a trace of crescendo-diminuendo. The introduction builds up rather romantically with long legato lines. The Allegro con spirito is taken steadily, but has a sprightly early morning air. At this tempo, van Remoortel makes more of the recollection of the theme of the introduction towards the end of the development than most conductors – many seem to just breeze through hoping for the best. There are two editing errors, one resulting in a seven-bar cut (bars 135-141) shortly after this reminiscence of the introduction. The other follows the drumroll near the end of the movement, after which the first note of the theme in the lower instruments is missing. No repeat. The second movement is slowish, but van Remoortel concentrates on sprightly staccatos, with no attempt to dig in and discover a funeral march in it. He plays the two repeats in the first statement of the minor key theme, then the first repeat in the violin solo variation, omitting all others. An unusual feature is that he makes a rallentando and a pause at the end of each variation. No conductor heard so far has thought to do this and it might become tedious on repetition. The minuet is slow but without either the ceremonial weight or the grand lilt that might justify the tempo – it just slogs. The trio, in spite of careful phrasing, stagnates. The finale, again taken steadily, impresses with its feeling of hushed expectancy, but the forte episodes, when they come, are too tautly controlled to make full impact. Van Remoortel makes much of the long-held wind notes, to good effect. All the same, not a strong contender.  

Mogens Wöldike’s set of the last six Haydn symphonies for Vanguard made use of Robbins Landon’s researches. In the “Drumroll” the music as played corresponds to the Robbins Landon score – no expressive bulges or pointless accents and the small differences to the phrasing compared with old editions are scrupulously observed. After an opening timpani crash to awaken the dead, followed by a long diminuendo, the introduction is moderately slow. Wöldike does not attempt to build it up romantically, creating instead an atmosphere of glacial expectation, anticipating, perhaps, the “Chaos” music of “The Creation”. The Allegro con spirito is upfront and robust at a swift tempo, with the trumpets allowed to shine through at climaxes. The second movement has a sort of subdued perkiness. There is no attempt to make a funeral march of it, or a triumphal entry with the C major theme, but the graver emotions are not passed over. The Minuet is bouncy with a graciously flowing trio. The finale begins with repressed excitement that eventually explodes in the forte passages. I have to say, though, that apart from the edition used, the basic concept is not so very different from Blech’s live performance, even if Wöldike goes that little bit further down the same road. That is to say, no-holds-barred Haydn that does not attempt to make Haydn sound like anyone else. As to repeats, we have Anthony Hodgson’s word[6] that Wöldike made all of them, but some were cut by the engineers to fit each symphony on an LP side. In this case, the first movement has its repeat, the second all but those of the C major section on its first appearance. I do not know if any CD issue has gone back to the original masters to reinstate these repeats.

Eugen Jochum’s 1958 recording begins with a controlled crescendo-diminuendo drumroll, but the phrasing and dynamics suggest he is using an Urtext score. His introduction is slow, but eloquent phrasing prevents it from being ponderous. If this suggests he is going to offer a romanticized interpretation, the Allegro con spirito, if not especially fast, has a joyful spring. The conductor’s romantic tastes cause him to make a slight rallentando into the second subject, which then skips along in tempo. The repeat is taken. His second movement is similar to Wöldike’s, a gentle march with suggestions of pastoral serenading when the C major theme comes round for the last time. All repeats are played – if not the first time since Golschmann, the first time the engineers left them in. The result is a 32-minute side, which was long in those days. Jochum’s minuet is a delight. Somehow he has managed to lighten the texture so it comes across with a sort of gawky elegance. There is exquisite shading in the string phrases after the double bar and a graceful poise to the trio. The finale is quite swift, but rather than sounding driven, its clear textures beam enjoyment by all concerned. A lovely performance that turns into a wonderful one for the last two movements.

Back in 1943, Jochum’s live Hamburg recording, issued on LP by Melodiya and more recently on CD by Dante Lys, offers a glimpse of the Furtwängler performance we shall never hear. After a very gentle, distant drumroll, the introduction rises from the morning mists, contemplative rather than ominous. The Allegro con spirito is taken at a fairly relaxed tempo, with some tightening in fortes and a slight holding back in piano sections here and there, such as that heralding the recapitulation. It must be said that Jochum’s Haydn was not heavy-footed even in his romantic youth. The repeat is played. The second movement is closer to Heger’s funeral march conception, with wailing winds, but also with a tendency to move on in forte passages, especially when they are in the major key. Three repeats are omitted. Jochum’s minuet was fairly spritely even in 1943 and he shapes the string phrases after the double bar very tenderly, but whereas in 1958 he managed this without losing tempo, fifteen years earlier he lingers unduly over them. The trio is nicely shaped without the special quality he found later. The finale has plenty of adrenalin, and a few spurts of faster tempo as the fortes burst out. I have to say that, thanks to the conductor’s obvious involvement, I did not find the occasional tempo vagaries disturbing, though they might loom larger on repeated hearings.

Beecham’s second recording (ASD 341) seems to be a self-conscious and counter-productive attempt to improve on his first. After a small crescendo-diminuendo drumroll, the introduction is slow and portentous. The Allegro con spirito is steady, with an insistent brightness, though the second subject is buoyant. Still no repeat. His second movement allows one more repeat than before. As well as the first repeat, he plays the repeat in the C minor section following the solo violin variation. His tempo is almost didactically slow. He can just about hold the attention, but it is too much of a public monument. His minuet and trio get all their repeats this time. The minuet skips along and it is certainly not heavy but it becomes over-insistent. The trio is nicely shaped. The finale begins and remains steady, with an almost pedantically strong accent on the first beat of every bar. The excitement of the earlier version is never approached. A disappointment.

I will discuss Dorati’s 1958 recording in conjunction with his performance in the complete cycle.

Revered at the end of his career, back in 1959 Günter Wand was working quietly in Cologne, known to discophiles only for a handful of recordings for the Club Français du Disque, of which his “Drumroll” was one. Could those few that bought it have guessed this was a conductor who would one day be hailed as great? If they had ears to hear, I think yes. The striking feature of this performance is, in a way, its lack of striking features, its complete avoidance of the sort of personalized phrasing or dynamics that draw attention to the interpreter. A non-interpretation, almost, yet one that somehow holds the attention, directing the ear to the music and only the music and leaving the aftertaste of a memorable experience. The sort of thing Klemperer could do at his best. Wand begins with a small crescendo-diminuendo drumroll, after which the introduction unfolds as a spacious contemplation. The Allegro con spirito is relatively slow, but buoyant and alert. No repeat. The second movement has a midway tempo and avoids the spiky staccatos of Beecham’s 1958 version. The steady tread suggests a pilgrims’ march. Wand takes the first repeat of each section, omitting the second. This is at least consistent. The minuet goes at about Beecham’s 1958 tempo, but without his over-insistence. Buoyant is again the word. The trio is a good example of how Wand’s apparently straight approach draws the ear into the music. The finale is steady but without Beecham’s jabbing accents on the first beat of each bar. It does not thrill, but as it continues it generates a sense of wonder. A remarkable performance.  

Remarkable in another way is Igor Markevitch’s recording with the Lamoureux Orchestra, of which he was principal conductor from 1957 to 1961 and which he evidently trained to a very high standard (Philips A 02041 L). After a fairly contained crescendo-diminuendo drumroll, the introduction has a brooding quality, as though waiting for some great event to burst out. The Allegro con spirito is very fast, contrasting feather-light scherzando pianos with eruptive fortes that remind us the explosive force for which Markevitch’s Rite of Spring was renowned – though with clear, luminous textures it still sounds like Haydn. No repeat. Markevitch – who acquired Italian nationality in 1948 and was very active in that country – perhaps knew better than most conductors what Andante più tosto allegretto really meant. He takes the movement at a forward-moving two-in-a-bar, but with such balletic lightness in the piano sections as to avoid any sense of hurry or rigidity. There is a joyful spring to the major sections. In the violin solo variation, the player meets the challenge of the fast tempo imperturbably. As with Wand, there is a feeling of a pilgrims’ march, but a less penitential one. That Markevitch may have had some sort of programme at the back of his mind is evident when, before the coda, he makes a ritardando, as though to say “we’ve got there”, then shapes the pianissimo bars (167-170) so exquisitely that the record would be worth having only for that. Another interesting feature is that in the final big statement of the C major theme, where Haydn has underpinned the music with repeated semiquavers on the timpani, Markevitch plays exactly this. Robbins Landon has indicated here in editorial square brackets a trill sign. He clearly supposed, as most of us would, that Haydn expected a drumroll, whereas with Markevitch, and thanks to the very clear recording of the timpani, we hear the throbbing of the distinct repeated semiquavers. That is what Haydn wrote and Markevitch proves it sounds effective.  Just two repeats are played – the first and the first in the violin solo variation. The minuet has a ceremonial feel, but at a faster tempo than the ceremonial performances heard earlier on – almost like a grand ceremonial ball. We only realize how fast it is when the trio literally scuds along. The finale keeps up a remorseless rhythmic impetus, finding space to phrase the piano sections with lightness and grace and to give the fortes eruptive force while keeping everything luminous and clear. If the recording were not so good, you could fob this off as a performance from Toscanini’s best years.

Bernhard Paumgartner’s recording with the Camerata Academica des Salzburger Mozarteums was issued in 1960 (Opera 1150). Paumgartner (1887-1971) was born in Vienna but for most of his career (from 1917 onwards) was associated in some way with Salzburg and particularly the Salzburg Mozarteum. The orchestra is not the Salzburg Mozarteum Orchestra itself but an offshoot, today called simply the Camerata Salzburg, which Paumgartner founded in 1952. This appears to be the first time the symphony was recorded by a chamber orchestra other than an English one. After a gentle drumroll, Paumgartner takes the introduction very slowly, with no attempt to bind the phrases together. The Allegro con spirito is taken steadily though it has a certain sprightliness and is not heavy. No repeat. The second movement clunks along amiably, whether the key is minor or major. There is no apparent attempt to go beyond the notes. Just two repeats are played – the first, and that of the C minor section following the violin solo variation. The minuet has a certain elegance while the trio has an agreeable flow. Both repeats are omitted from the minuet on its first appearance, though the trio is played in full. I suspect these repeats were removed by the engineers out of desperation – at Paumgartner’s slow tempi, the symphony lasts 28½ minutes even with repeats cut to a minimum (so perhaps he took other repeats too). I would describe this performance as provincial (even though Salzburg is theoretically a major city) since its limitations are not so much those of the performers – as might be the case with van Remoortel’s recording – but arise from a blinkered unawareness that, in the big wide world outside Salzburg and its Mozarteum, bigger and bolder things happen.

Horst Stein’s Dresden version is one that I was unable to track down.

Paumgartner was one of Herbert von Karajan’s teachers at the Salzburg Mozarteum – for composition and chamber music, not conducting. Whatever charges you wish to lay at Karajan’s door, provincialism is not one of them, yet it is curious to what extent his 1963 Vienna PO recording (Decca SXL 6067) comes across as “corrected Paumgartner”. After a mistily distant drumroll, his introduction is consistently subdued, with no apparent attempt either to build it up or to bind the phrases together. Yet, by some sort of personal magnetism, he can hold the attention with this approach where Paumgartner could not. His Allegro con spirito lollops along nicely, beginning almost conversationally then breaking into sprightly fortes which, though the full VPO seems to be in attendance, do not become heavy. The tempo is a bit fresher than Paumgartner’s and the repeat is taken. His second movement is within measuring distance of an Allegretto and plinks along nicely, without trying to mine hidden depths – the wailing wind later on sound agreeable, no more. He omits second repeats in the C major section, both the first time round and in the violin solo variation. His minuet is fairly slow and formal. He succeeds in avoiding heaviness or a sense of the ceremonial, but I came to find it laboured. The trio is polite. His finale, though a little brisker than Paumgartner’s, bumbles along pleasantly. The striking thing about all this is, given the great name on the podium, how free the performance is from point making. I am left puzzling as to how Wand’s apparent non-interpretation has the stuff of genius whereas Karajan’s is a well-behaved performance that, if it does not diminish the symphony as Paumgartner’s risked doing, does it no great favours either.

Since conductors did not exist in Haydn’s day, it was time one of the smaller chamber orchestras essayed a version without a conductor. The Prague Chamber Orchestra obliged in 1967 (Supraphon 1 10 0220). Mike Gray’s catalogue and some other sources say that the performance is conducted by Jiří Ptáčník even though the record label says clearly “bez dirigenta” (without conductor) and describes Ptáčník as “Umelěcký vedoucí” (Artistic Director). I suppose this means he leads from the violin, but it might equally mean he just sat there issuing commands. By the time of the London Symphonies, Haydn himself sat at the fortepiano, but apparently controlled the performance more by gesticulating where needed and did not necessarily play all the time. So I suppose you might describe him as an Artistic Director. The performance begins with a long drumroll and a controlled crescendo-diminuendo. The introduction is very slow and somewhat cautious, but just as you think it is getting nowhere, the players burst into a passionate forte on the second beat of bar 29. There is no such marking in the Robbins Landon score and presumably not even in the old corrupt editions, since no performance I have heard so far has thought to do this. It is highly effective, but Haydn, though sparing with dynamics, unfailingly marked pianos and fortes where he wanted them. The Allegro con spirito is very fast and skips along delightfully – an effect reinforced by playing acciaccaturas not appoggiaturas in the second subject. After the return of the introduction, the coda is played even faster than before, sounding like a race to the finish. No repeat. The second movement is taken rather slowly and gravely, though without making a funeral march of it. The tempo tends to move on a bit in the major sections, where the rustic Czech wind which so delighted collectors of early Supraphon LPs make an attractive effect. When the solo violin variation comes along, the player has evidently had enough of this trudge and plays the whole section in a faster tempo. Maybe this is where a little shouting and screaming from the rostrum would not have come amiss. Not a single repeat is played. The minuet is played quite slowly, but by making the three beats not quite equal, the players give it a rather folksy lilt. The interlocking lines of the trio show the chamber music style of performance at its best. No second repeats in either the minuet or the trio. The finale is a total success, not least because Haydn provided them with no repeats to omit. The tempo is very fast and scampers along infectiously, with the burbling Czech bassoon making some agreeable contributions. The opening four-bar theme for just two horns is played with the notes well separated – a plausible and effective idea, though every other performance heard till now has played them legato.

The cycle of “London” Symphonies by Leslie Jones and the Little Orchestra of London (Nonesuch HF-73019) aroused a good deal of interest in its day. Notable was the inclusion of a harpsichord continuo (played by Harold Lester) though, as far as no. 103 is concerned, I can only say that at one single point in the second movement I fancied I heard a slight ting from the instrument. Those who are happy enough to have a harpsichord included provided it cannot actually be heard may set their hearts at rest. In any case, by the time of these late symphonies, Haydn directed from the fortepiano not the harpsichord, and it is not clear how much he really played as long as things were holding together satisfactorily. Perhaps surprisingly, Jones leads off with a very gentle drumroll – just a slight crescendo-diminuendo. The introduction is fairly slow and does not attempt to impose any particular character on the music, merely letting it unfold as it is. The Allegro con spirito is taken steadily but with considerable vitality. The repeat is taken. I do not know whether the small dynamic range is due to the players or the recording, I suspect a bit of both. This is more problematic in the second movement. Jones takes a definitely Allegretto tempo, allowing the music to clip along agreeably whether in minor or major mode. There is no hint of pathos or indeed of any sort of expression. Maybe Jones felt such things were 19^th century accretions to be avoided and doubtless this could be refreshing after earlier attempts to equate the music to the “Eroica” funeral march. However, Markevitch in particular showed that the music can still say a good deal at a tempo not far short of this. Moreover, while this is pretty enough in a foot-tapping way at the beginning, as the variations become more complicated and contain more notes, there is a suggestion that the music is being frogmarched along for fear of slackening the pace. All repeats are played. The minuet is quite grand and formal, but just manages not to slog. In the trio, it is interesting to hear the wind instruments, which double the strings for much of the time but are often barely audible, brought forward as equal partners. The finale is steady but gains steam as it proceeds and unlike the steady performances of Karajan and Maazel, succeeds in rising to a convincing conclusion. It is with some relief that I read this comment from the knowledgeable Roger Fiske in his generally favourable review of the “London” symphony box. “I have not yet got used to Mr. Jones’s tempo of 103/2. Haydn originally marked it Andante, but then got cold feet that it might be taken too slowly, so he added a request that it should be played nearly allegretto. Mr. Jones can reasonably argue that this is what he has done. And yet somehow at this tempo the movement loses its stature, and it does not sound the great music one expects”^[7]. As for the other three movements, this is a fresh, alert account that does not set its sights very high. Wand had proved not long before that apparent non-interpretation can be raised to the level of genius and produce illuminating and inspiring results, but you do need that little spot of genius to bring it off.          

Lorin Maazel’s small group of recording for the Concert Hall Record Club, with the Berlin Radio Symphony orchestra, has not circulated widely. Nor did Haydn ever figure extensively in his programmes, though the omens were good since his Mozart could be excellent (SMS 2616). After a distant drumroll followed by a pause, the introduction is fairly mobile and manages to express a sense of unease. The Allegro con spirito is bright and lively at a moderately fast tempo. No repeat. The second movement is taken slowly, but rather than create a funeral march, Maazel concentrates on highly expressive phrasing. I found this most affecting. However, he needs to move on a little in the major sections, which assume a march character. Both repeats are taken in the first statement of the C minor theme, also the repeat on its return after the violin solo variation. The other repeats are ignored. The minuet is slow. With lean orchestral textures and sharp accents, he does his best to justify his tempo, but by the end I found it slogged while the trio, though carefully phrased, hung fire. The finale, like Karajan’s, chugs along nicely at a steady tempo without concluding anything in particular. Anyone to whom you played the C minor sections of the second movement might suppose this was one of the finest versions of all. Unfortunately, the impression is that these did indeed strike a chord with the conductor and, since he could hardly omit the rest of the symphony, he despatches it with generic energy and decent style.

Having been thoroughly disappointed by Leonard Bernstein’s live performance of Symphony 49, his “Drumroll” was a pleasant surprise indeed (CBC M-33531). His opening timpani roll goes off like gunshot – so far only Wöldike has favoured this solution. His introduction is among the fastest – sufficiently so to imply a degree of underlying urgency. His Allegro con spirito is then on the slow side, almost a quick six-in-a-bar, but with bristling vitality. He then relaxes delightfully – in mood, not in tempo – for the second subject. The repeat is played. His second movement is swift enough to count as two-in-a-bar and with the staccatos not too spiky – he consistently plays the first two notes of the theme with an unmarked slur. The music has a questioning air while the C major theme, not too brazen, comes as sunny reassurance. He allows the solo violin the mordents indicated in a footnote to the Robbins Landon edition. Drama is introduced later on. In bar 179, he has the phrase played by the flute rather than the oboe. Once again, he has clearly studied the Robbins Landon score, which shows that this is an alternative found in one of the sources. He plays all first half repeats, no second half ones – the logical solution if you do not wish to play them all – or if your engineers insist there is no space to do so. Bernstein’s minuet is superficially very similar to Maazel’s, leaving me to puzzle over why, far from slogging, it had me nodding my head to the rhythm. The trio does not languish either, coming across as a serene episode in the Elysian fields. His finale, too, goes at about the same tempo as Karajan’s and Maazel’s and yet, unlike theirs, it builds up to a grand conclusion. It must be said, too, that in spite of the fullness of the orchestra, Bernstein is able to prevent it from sounding like a later composer, but I would say this of Maazel too, maybe even Karajan, so I can only conclude that Bernstein’s extra engagement with the music carries the listener with him. I think I prefer faster tempi in all but the second movement, but this is a marvellous performance all the same. 

In the much-visited symphonies, do the complete cycles of the early 1970s have as much to offer as they do where previous versions had been few and far between? Well, Märzendorfer certainly does. He starts deceptively, with a traditionally distant drumroll, though the crescendo-diminuendo is quite considerable. This is deceptive. When the introductory music comes back before the coda, he has his drummer give an almighty crash. This really has the power to shock. The introduction itself is slowish with a dark, brooding atmosphere. The Allegro con spirito is a lively, upfront dance. With hard timpani sticks, ringing trumpets and (excessively?) forward wind, including rustic, almost rasping oboes, the music sounds freshly minted. The repeat is played. The second movement is almost as brisk as Jones’s, but Märzendorfer manages to find expressivity and pathos as well as considerable dynamic range. He also allows the music to breathe even within the tempo. There is a piquant, village festival feeling to the C major theme while the dramatic moments rage suitably. All repeats are played. The minuet is a one-in-a-bar rustic romp. Purists will shake their heads when the trio arrives. Introducing it with a schmaltzy Viennese upbeat, Märzendorfer takes it at a much slower tempo. The bucolic clarinets are allowed to dominate over the strings. Furthermore, they are made to phrase in pairs of two notes – not how it is written – sounding rather like a hurdy-gurdy. Fascinating. The finale goes at a great rate – I haven’t heard such a rip-roaring version as this since the 1950 Beecham. He plays Haydn’s original – in more ways than one – ending, jettisoned by the composer for publication but included by Robbins Landon as an appendix. Unfortunately, a few false notes from the trumpets have been left in. The producer’s haste to get the cycle finished and in the shops before the Dorati maybe took its toll^[8]. Still, for sheer open air exhilaration and a challenging rethink of what Haydn should sound like, this is on its own.

In the case of Dorati, I am not so sure, though he begins well with a forte crash from the drum. The introduction, taken quite slowly, has an air of expectancy. The Allegro con spirito, with the repeat, is pretty swift, even breathless. Dorati seems to be using hard timpani sticks, though he does not let his trumpets and drums ring out as Märzendorfer does. Still, so far so vital. His second movement goes at something like Märzendorfer’s speed. While Dorati avoids the sensation of pushing on at all costs, as with Jones, he does not attempt to express anything beyond an amiable serenade. When the more dramatic variations arrive, they seem unmotivated in such a context. All repeats are played. The minuet is formal, sufficiently well sprung to avoid heaviness, but there seems little attempt to make the music interesting here or in the trio. The finale is steady, though by differentiating between forte and fortissimo and with the help of the hard-sticked timpani, it does arrive at some sort of conclusion. This performance does have some good things, but none that have not been equalled or bettered elsewhere. If Märzendorfer is too radical for you and you want all the repeats, Jochum’s 1963 recording is the place to go. 

As I was somewhat underwhelmed with Dorati, I was tempted to pass on his previous version from 1958 (Mercury SR 90208), also with the Philharmonia Hungarica. However, reliable commentators have suggested that his sporadic earlier forays into Haydn were much better, so duty called. He starts with a distant drumroll and the introduction here has a groping air. Like Wöldike, he is perhaps seeking a parallel with the “Chaos” music that opens “The Creation”. His Allegro con spirito back then was very steady – his repeatless performance is not much shorter than the 1972 traversal with the repeat. There is, however, a good deal of rhythmic kick and he makes quite a lot of the hint of the theme from the introduction towards the end of the development. His second movement had only three repeats in 1958 – the first two and then the dramatic C minor variation following the violin solo section. More than a serenade, he gives the music a military clip, not a funeral march but the ominous, steady tread of soldiers approaching and finally entering the city. Not unlikely thoughts for a Hungarian in 1958. The minuet is much as it was to be later, but the finale has that little bit more flair – an alert, wide-awake air that is rather attractive. So, in spite of the shortage of repeats, the earlier performance is preferable in three movements out of four.  

Turning now to versions that are either live or studio recordings for radio broadcast, Jochum’s 1943 has been discussed in comparison with his 1963 version. Hermann Abendroth’s recording with the Leipzig Radio Symphony Orchestra tends to confirm that senior German conductors of the mid-century saw this symphony as a precursor of Beethoven in E flat major. The opening drumroll begins piano and has a fairly contained crescendo-diminuendo. The introduction is not too slow and concentrates on the long line. The Allegro con spirito is an ebullient, Eroica-scherzo-related affair, with a nicely shaped second subject. No repeat. The second movement adopts a funeral march mould, if not quite as grave as Heger’s. The C major music has a heroic air on its first and last appearance. The solo violin is not very sensitive in his variation. All first parts have their repeats, second part repeats are omitted from all variations. More than ceremonious, the minuet is jubilant, with a good lilt and not too slow. The trio is unusually robust. Jubilation continues in the finale, which is taken swiftly with a good deal of flair. Though strong and energetic, this performance goes less far in the Beethovenian direction than Heger’s, relating as much, perhaps, to Mozart in E flat as to Beethoven.

Hans Schmidt-Isserstedt comes closer than anyone to replicating Heward’s pre-Schubertian concept. The drumroll is kept to a minimum, followed by a fairly mobile introduction concentrating on the longer line. The Allegro con spirito dances along delightfully, with a slight slackening for the second subject. The repeat is played. The second movement is slowish but, as with Heward, suggests a Schubertian amble through the countryside, tinged with sterner thoughts from time to time. The violin solo variation soars like a bird through the sky. Schmidt-Isserstedt plays just the first repeat and the first repeat of the violin solo variation. He also omits the second repeat of the minuet. This is taken fairly slowly but with a nice lilt and a very well-phrased trio – Schmidt-Isserstedt finds more in this than many conductors. The finale spins along joyfully. A lovely performance, if short on repeats.

I have discussed Blech’s live performance above. The Cameo Classics/Richard itter box also has a 1959 performance by Harry Newstone and the Haydn Orchestra. Newstone begins the drumroll piano but makes a big crescendo-diminuendo. The introduction unfolds with brooding deliberation, breaking into a very lively Allegro con spirito. There is abundant vitality here and the second subject skips along without any slackening of the pace. I have to say, though, that the playing in the forte passages is a little crude. His second movement moves at a good tempo, encompassing both the doleful first theme in the minor key and the more perky second theme in the major. There is no especial distinction to the phrasing, though. Newstone omits both repeats in the first statement of the major key theme but plays all others. There is not much lift to his grandiose minuet and it comes across as rather heavily scored. His finale is full of vital energy, the pervasive repeated notes clearly looking ahead to Beethoven’s Fifth Symphony. Unfortunately, the sometimes crude playing anticipates some of the more tub-thumping performances of that work. I found this disappointing – all the other Newstone performances I have heard reached a uniformly high standard, even when surpassed by other versions. Out of fairness, I will mention that my colleague Stephen Greenbank found Blech and Newstone “equally captivating” and was “hard pressed to choose between the two”^[9].

Outside France, Manuel Rosenthal (1904-2003) remains obstinately associated with his Offenbach ballet “Gaîté Parisienne” but at home he was the composer of many serious works and the conductor of a wide range of music – including Vaughan Williams’s Fifth Symphony, a performance I would love to hear if it survives. He showed sporadic interest in Haydn – not just the last twelve. His 103 begins with a distant drumroll but builds up considerable tension during the introduction – enough to silence the coughers and shufflers who greeted the first page. The Allegro con spirito is high spirited indeed, and Rosenthal’s operatic experience enables him to ease into the second subject without actually losing tempo. The repeat is played. The second movement gets only its first repeat – a rather jaunty tour through the music. If the big rallentando at the end of the solo violin variation was intended to prepare for the dramatic interruption that follows, I do not think this works. The medium-paced minuet has enough lift to keep it afloat, but no great distinction, here or in the trio. Best is the finale where, though the playing is not very refined, Rosenthal generates real electricity. Enjoyable enough if you were there but there are better ways to remember this conductor.

Lovro von Matačić (1899-1985) was a frequent guest with the Italian radio orchestras and invariably obtained vital and cohesive playing from them. His Naples performance of Symphony 103, though using a chamber orchestra – the Naples Scarlatti Orchestra was originally a sort of Italian parallel to the London Mozart Players – is anything but small-scale. He starts with a distant drumroll, without too much of a crescendo-diminuendo, followed by a pretty slow introduction that invokes a sense of numbed pain. The Allegro con spirito is not all that fast, but with plenty of energy in the quavers it expresses a peremptory vigour, while characterizing the second subject with elegance and cunningly pointing the harmonic bypaths of the development. The repeat is played. The second movement suggests a funeral march, but with very short staccatos it avoids suggesting that of the “Eroica”. Unobtrusive conductorial control is shown in the way, after the first double bar, Matačić minutely delays the upbeats to obtain a sighing effect from the drooping phrases in the following bars. He plays all first repeats in each variation, none of the second ones. Unlike any other conductor I have heard, he has the solo violin (probably Giuseppe Prencipe, a fine artist) play the first repeat differently – gently and sweetly the first time round, then a suddenly almost aggressive attack as the repeat begins. The minuet is not so very fast, but has a peasant-stomping outdoor quality, while the conductor’s detailed phrasing coaxes more character from the sometime inert trio than many conductors achieve. The swift and fiery finale, carrying all before it, caps a splendid performance. Anyone thinking of issuing a survey of Matačić’s RAI recordings – it would be a very good idea – should not overlook this.   

Jean Martinon’s performance comes from fairly late in his not entirely happy period with the Chicago Symphony Orchestra – though the unhappiness seems to have derived from an all-powerful critic rather than any shortcomings of his own. He begins with a distant drumroll and an introduction that is very long-drawn indeed – only a few touches of warmth counteract the icy stillness. The Allegro con spirito is not especially fast but has considerable vitality. No repeat. The second movement is very slow, but the effect is not so much a funeral march as a personal, mainly forlorn meditation. I thought the opening not piano enough, but this proves to have been deliberate – the repeat is played with a real pianissimo. The whole movement is notable for its dynamic shading. Typically, after the first double bar, two bars are played mezzo piano, followed by a sudden drop to pianissimo and then a carefully controlled crescendo. Only the first two repeats are played. The minuet is a pretty swift, one-in-a-bar affair, with a slightly slower tempo for the neatly phrased trio. Up until this point, I found the interpretation interesting and sometimes unusual, but with a certain reined-in feeling that perhaps stemmed from the uneasy relationship between the conductor, the orchestra and the Chicago public. The finale succeeds in overcoming this. It does not seem so very fast at the beginning, but it acquires sizzling vitality as Martinon finally succeeds in firing up the orchestra. I should be interested to hear Martinon conducting Haydn with an orchestra with which he had an easier relationship. 

Those who investigate Dean Dixon (1915-1976) on the strength of his meagre official discography find it difficult to account for the gap between the firebrand of the early Westminster/Nixa recordings and the grandly expansive, seemingly uninflected recordings made in Prague in his last years, and are inclined to assume that he simply lost his spark. While it is true that Dixon’s odyssey through the world of music is poorly documented through commercial records, most of his European appointments were with radio orchestras and, though the broadcasting stations have not preserved everything, there would be enough to convert him from one of the least-recorded to one of the most-recorded conductors overnight if anyone were to share my opinion that it was worth doing. Since I wrote about Dixon for MWI more than a decade ago^[10], a great deal of material has come my way, though most of it circulates only privately, including his Haydn 103. He begins with a distant drumroll, after which the introduction is slow and brooding, while gradually generating expectancy. The Allegro con spirito is unhurried, but energized from within and not heavy. The repeat is taken. The key to this performance is perhaps the second movement. The tempo is broad, yet Dixon manages to avoid the suggestion of a funeral march, whether tragic or heroic, by concentrating purely on musical values. The notes tell their own tale, yet there is an overall conductorial control that prevents it from plodding and holds the attention. The leading violinist plays his solo with a gentle lyricism. All repeats are played – only Golschmann had dared to include all repeats in a slow tempo. The minuet is big boned, the trio very well phrased. Dixon’s pace in Haydn’s finale is not unduly fast, yet there is fire as well as grandeur, concluding a performance that remains in the mind for its big-hearted conviction. The obvious parallel is Klemperer – who did not record this symphony – and Dixon demonstrated his admiration for Klemperer by conducting his Second Symphony. Perhaps if he had lived and remained active another couple of decades he would have assumed a mantle similar to Wand’s – Bruckner was another composer he did exceptionally well.

Conclusions? Märzendorfer marks a sort of springboard between old and new – a bold approach presaging some of the HIP developments that were to come. As noted, though, there are some oddities and spots of rough and ready playing. Much of his boldness is also found in Wöldike, especially if you can find it with the missing repeats reinstated. From Jochum you get all repeats in a more traditional yet wonderfully fresh performance. If repeats are a must and you want a slow tempo for the second movement, it is back to the 1935 Golschmann or you can hope for an official issue of the Dixon. Shorter on repeats, several performances offer special qualities – Heger for his demonstration that this symphony can convincingly look ahead to the “Eroica”, Heward (or Schmidt-Isserstedt) for finding pre-echoes of Schubert. Remarkable performances in their various ways come from Beecham (1951), Markevitch, Wand and Bernstein, and I would add the live Blech and the Matačić to these. Plus a thought for Martinon. Moreover, freshness and vitality of the music itself invariably won the day even in the performances I found less successful.

Some interesting things emerge from the following table – for example that Leslie Jones takes slightly less time to play the second movement with every repeat than the Prague Chamber Orchestra take to play it with none at all. It also induces some thoughts on the difference between perceived tempi (as described above) and actual tempi as per the stopwatch. Evidently phrasing, texture and general characterization of the music can make us hear it as faster or slower. Given that it is unusual to play a minuet and trio without repeats (except in the da capo of the minuet), it can be assumed that these are played unless indicated otherwise.

	I	II	III	IV
Golschmann	07:15 with repeat	12.08 with all repeats	05:10	05:03
Heward	08:28 with repeat	08:07 5 repeats omitted	04:00	04:55
Solti 1949	07:58 with repeat	07:52 3 repeats omitted	04:21 2nd repeat of trio omitted	04:38
Heger	09:47 with repeat	11:01 2 repeats omitted	05:16	06:02
Scherchen	10:00 without repeat	10:41 5 repeats omitted	05:07	05:27
Munch	08:30 with repeat	07:44 6 repeats omitted	05:03	04:56
Beecham 1951	08:18 without repeat	09:04 6 repeats omitted	03:22 2nd repeats omitted from minuet & trio	04:51
Blech 1955	08:12 with repeat	09:11 5 repeats omitted	04:28	04:59
Blech 1956	08:23 with repeat	09:25 5 repeats omitted	04:18	05:09
Van Remoortel	08:08 without repeat	10:02 4 repeats omitted	05:38	05:29
Wöldike	09:17 with repeat	10:23 2 repeats omitted	04:35	05:28
Jochum 1958	10:19 with repeat	10:33 with all repeats	05:11	05:11
Jochum 1943	10:31 with repeat	09:54 3 repeats omitted	04:56	04:57
Beecham 1958	08:47 without repeat	10:58 5 repeats omitted	04:54	05:34
Wand	08:46 without repeat	09:21 3 repeats omitted	05:05	05:19
Markevitch	07:34 without repeat	07:57 5 repeats omitted	04:47	05:14
Paumgartner	09:50 without repeat	08:34 5 repeats omitted	03:51 both repeats omitted from minuet	06:16
Karajan VPO	09:51 with repeat	09:20 2 repeats omitted	05:28	05:11
Prague CO	08:14 without repeat	08:58 all repeats omitted	03:46 2nd repeats omitted from minuet & trio	05:01
L. Jones	10:08 with repeat	08:49 with all repeats	05:02	05:35
Maazel	07:48 without repeat	10:08 4 repeats omitted	05:49	05:40
Bernstein	09:32 with repeat	09:08 3 repeats omitted	05:37	05:45
Märzendorfer	09:29 with repeat	09:23 with all repeats	04:57	05:08
Dorati 1972	08:47 with repeat	09:37 with all repeats	05:17	05:24
Dorati 1958	08:09 without repeat	08:31 4 repeats omitted	05:18	05:13
Abendroth	07:39 without repeat	09:59 3 repeats omitted	04:27	05:09
Schmidt-Isserstedt	09:23 with repeat	08:29 5 repeats omitted	04:31 2nd repeat of minuet omitted	05:05
Newstone	09:37 with repeat	09:15 2 repeats omitted	04:49	04:54
Rosenthal	09:52 with repeat	07:30 6 repeats omitted	05:09	05:12
Matačić	09:56 with repeat	08:52 3 repeats omitted	04:45	05:09
Martinon	08:51 without repeat	09:15 5 repeats omitted	04:41	05:01
Dixon	10:04 with repeat	12:04 with all repeats	05:52	05:18

© Christopher Howell 2026

---

[1] It is beyond the scope of this article to discuss whether Brahms’s own Andante movements should really be played as Adagios.

[2] Clough, Cuming: The World’s Encyclopedia of Recorded Music, Sidgwick & Jackson Limited, 1952; ditto, Second Supplement, 1953; ditto, Third Supplement, 1957

[3] https://charm.rhul.ac.uk/index.html

[4] https://classical-discography.org/

[5] See also my 2915 article on Golschmann: https://www.musicweb-international.com/classrev/2015/Jul/Golschmann_forgotten.htm

[6] Hodgson: The Haydn Seekers, CRC Winter 2001

[7] Gramophone March 1969, p.1286.

[8] Kurt List, the producer, told Anthony Hodgson that he would not have wanted “Dorati to have gotten the jump on us” (CRQ Magazine, Spring 2014

[9] https://www.musicweb-international.com/classrev/2020/Dec/Haydn-sys-CC9119.htm

[10] https://www.musicweb-international.com/classrev/2014/Jan14/Dixon_forgotten.htm