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. 29 in E major Hob. I/29 (1765)

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

The most striking feature of Symphony 29 is its extreme sparseness. For much of the first, third and fourth movements, the first and second violins play the same music while, almost as often, the violas double the cellos and basses an octave up. The first theme of the first movement has a basically long melody divided into questions and answers between the strings and the oboes which, unlike the strings, play in pastoral thirds. Thereafter, neither the oboes nor the horns play a very prominent role and have surprisingly few opportunities to warm up the predominantly two-part writing. In the second movement, the phrases of the melody are broken up so as to pass continually from the first violins and the seconds, and then back again. Remember that, in the “good old days”, firsts and seconds sat on opposite sides of the orchestra, so this must have sounded far more curious and fascinating to Haydn’s first audiences than we can imagine today. Only at the end of the movement do they come together to play the tune complete in octaves. The sparseness reaches its apogee, however, in the trio of the third movement. The music suddenly shifts to E minor and, in Antony Hodgson’s words, “Over an oompah accompaniment in the strings, the horns play [a long held note]. But what about the tune? Unless the harpsichordist provides one it simply has to remain absent, and no keyboard player has dared to provide such a thing for a quarter of a century”[1].  Implicitly, Hodgson seems to think this should be done. In a footnote, he tells us that “The last time was in Loibner’s recording when Christa Furhmann gave the outline of a melody”. Robbins Landon, on the other hand, felt the strange effect was intended, praising its “sombre and secretive beauty”: “The weird atmosphere of this little E minor trio is enhanced by the dark-hued pedal point in the horns and the total absence of any melody whatever”[2]. Another point made by Robbins Landon is that the most powerful movement is the finale – normally, at that date, finales were light and entertaining even when the first movement was strong and serious. “Here is Haydn at his most aggressive and brilliant. There is only one small episode in piano”[3]. Provided, obviously, that conductors do not add dynamic markings of their own – the chromatic passage from bar 54 offering particular temptation.

The first recording was the above-mentioned version in which Wilhelm Loibner conducted the Vienna Academy Chamber Orchestra. This was issued in 1952 on Lyrichord (LL36). It is particularly frustrating that I have been unable to hear this in view of Loibner’s – and Furhmann’s – hypothetical realization of the trio. I should be interested to know what it sounded like. I understand some more recent recordings have followed this example, but within my cut-off date, the only other recordings are that by Leslie Jones and the Little Orchestra of London (Nonesuch H-71121, issued in 1966 in the USA), and those of the two complete cycles by Märzendorfer and Dorati. Märzendorfer did not habitually use the harpsichord in his recordings, nor does he here. A harpsichord is occasionally heard in Dorati’s recording, but not in the trio. The harpsichord is better integrated into the texture in the Jones version, in which a bassoon can also be enjoyed at times, especially in the first movement. But the trio, maybe in deference to Robbins Landon, is played as is.

Of these three, Jones emphasizes the closeness of the music to the baroque age. His first movement is amiable but robust, with a steady vitality. The violin triplets towards the end of each section are played with separate bows, not slurred as written. The result is undeniably more brilliant, but I have to ask, not for the first time in this series, why use an Urtext edition if you are going to edit it to your own ends? The first repeat is played.

While graciousness is not a frequent feature of Märzendorfer’s readings, his shaping of the principal theme in the first movement, with gentler tones than usual from his oboes, achieves just this. Vitality comes later, as the violins launch into triplets, which he and also Dorati play slurred, as written. Dorati’s approach is surprisingly similar, indeed – the only reason for preferring one or the other is that Märzendorfer plays both repeats while Dorati gives only the first.

A certain grace also informs Märzendorfer’s second movement, where Jones prefers a steady baroque-like tread. Dorati is more rigid still with the staccato bass line. It is in this movement that his otherwise evanescent harpsichord suddenly comes out of nowhere to add some curious little jangles over the forte octave interjection by the lower strings in bars 16-17. I do not see the point of this. Märzendorfer and Dorati give the first repeat only, Jones omits both.

Jones plays the minuet in a robust, strutting manner that becomes  a little too insistent as it proceeds. His trio is the most literal of all, but also the version which finds the best balance between the weird horn notes and the comic “oom-pah-pah” of the strings. Rather than seek poetry, Jones presents it in all its tuneless bareness, as if the Haydn joke is that all the time you expect a tune to start and none does.

Märzendorfer’s piercing oboe may seem slightly disconcerting at the beginning of the minuet but you could argue that Haydn, by starting the movement with such a high note for the oboe, must have expected something like this. If you disagree, Dorati keeps his player well under wraps, as did Jones. Apart from that, there is little to choose between them – this is not one of Dorati’s flabby minuets and both have a good, upfront swing. In the trio, Dorati makes rather more of the odd harmonies.

Which of the three you prefer in the finale will depend on whether you agree with Robbins Landon’s description of the music, quoted above. Not for the first time, if you want an illustration of Robbins Landon’s point you have to go, not to Dorati, even if Robbins Landon was acting as musicological supervisor and wrote the notes, but to Märzendorfer, or maybe Jones. Märzendorfer keeps at a steady forte, with little dynamic variation except for the sudden drop to piano where Haydn wrote this. He also makes both repeats, though an unclean violin attack at bar 45 occurring exactly the same way both times raises unkind speculation as to whether the music was actually played twice over. Jones seems relatively underpowered at the beginning, but he gains strength as he goes on, sticks to Haydn’s dynamics and plays both repeats. The result is a little unremitting from both conductors, Märzendorfer especially, but Robbins Landon implies that it is supposed to be so. Dorati adds a fair amount of unwritten dynamic shading and omits the second repeat. With vital playing, the result is rather more listener-friendly. Doubt remains as to whether it ought to be so.

On balance, the vote seems to be for Märzendorfer. Just one broadcast performance has come along to swell the ranks – as often, it is to be found in the Cameo Classics box deriving from the Itter Collection. This is from a concert given on 22 September 1952 in which Hermann Scherchen conducted the Royal Philharmonic Orchestra, a programme in which the symphony was sandwiched between Haydn’s Te Deum no. 2 and “Creation” Mass. At about this time, Scherchen recorded Haydn’s “London” Symphonies and a few others in Vienna, with invariably memorable, if controversial, results. Symphony 29 was not part of that recording programme, so this is a discovery of some significance – it would be nice to think the rest of the programme languishes in the Itter archive and may eventually see the light of day. The sound is good for the date.

Scherchen is a little broader than the other two in the first movement. It is worth mentioning that, though Haydn’s final marking was Allegro di molto, he originally wrote Allegro ma non troppo. I have not seen the old Breitkopf edition, which Scherchen presumably used, so I do not know how it is marked there. The music sounds fuller, less sparse, in Scherchen’s hands. I do not know how he achieved this, since he adds nothing that is not in the score, though I suppose he might have doubled the oboes and horns. There is a touch of sublimity in the way he unfolds the opening theme, and the slower tempo does not preclude vivacity when the triplets arrive. First repeat only. His second movement is pretty slow. If you were actually there, the effect of “visual stereophony” as the first and second violins passed their wisps of melody from one to the other probably justified the tempo and I must say he invests the forte passages with an almost Beethovenian power. First repeat only. His minuet is a grand, ceremonial affair, but he manages to keep it afloat and he really digs into the strange harmonies in the trio – we actually hear them better in this 1952 recording than in the other three from many years later – probably because of the way Scherchen balanced them. The finale is similar to the first movement – slightly on the slow side but with plenty of fullness and power. Like Dorati, he adds dynamic shading of his own and plays only the first repeat. In spite of a few quirks, I think this is the version I will return to more often. Scherchen manages to find hints of incipient greatness in this apparently unassuming early work.

Christopher Howell © 2026

[1] Antony Hodgson: The Music of Joseph Haydn: The Symphonies, Fairleigh Dickinson University Press 1976, pp. 65-66.

[2] H.C. Robbins Landon: Haydn. Chronicle and Works. Vol. 1. Haydn: the Early Years 1732-1765. Thames and Hudson 1980, p.571.

[3] Idem.