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. 34 in D minor Hob. I/34 (1765 or earlier)

Symphony no. 34 is one of several symphonies from the 1760s adopting the Sonata da Chiesa pattern with an extended slow movement placed first. The other movements – a terse Allegro, a Menuetto and Trio and a dance-like Presto assai are in the major key while retaining a tone of seriousness, even toughness. The symphony amounts to a fully coherent whole. If it is less popular than certain other symphonies that start with the slow movement, this must be due to the lack of a name. A curious feature of the trio is the syncopated horn parts accompanying the oboes carolling in thirds.

The earliest recording of this symphony, so far as I can discover, was set down in June 1949 by Jonathan Sternberg and the Vienna Symphony orchestra (HSLP 1002).[1] I have been unable to hear this.

Unknown until recently, a performance by Harry Newstone and the Haydn Orchestra has been issued in the Cameo Classics/Richard Itter Collection box of early Haydn performances (reviewed here). Newstone was born in Canada in 1921 and died there in 2006, but most of his career was based in the UK. He formed the Haydn Orchestra in 1949, a small band which, while using modern instruments, sought to develop a proper classical style. He later went to Italy to study with Fernando Previtali and became an effective conductor of a wide repertoire, with a penchant for neglected works – he premièred five of the Havergal Brian Symphonies. His small discography contains just two Haydn discs – the Haydn Society/Nixa coupling of Symphonies 49 and 73 (1952) and the Oiseau Lyre pairing of 46 and 52 (1959) – so the addition of three symphonies in the Cameo/Itter box almost doubles his meagre representation.[2] The performance of no. 34 is dated 23 November 1954 and described as live, but no audience is discernible and I suspect it was set down in the studio for broadcasting purposes.

In Newstone’s hands, the first movement unfolds with a breadth that has a touch of sublimity to it. He does not play either repeat. He gives the first repeat in the second movement and all the many repeats in the finale except, oddly, the last. There is strength and vigour in the second and fourth movements and a nice lilt in the third. A word may be needed about playing style. String players in those days expected to play fully every note, without gaps between unless specifically commanded. Even staccatos were not jabbed at but allowed to sound before leaving the note. So this performance, in spite of the reduced orchestra, has a fullness and a singing quality contrasting with the spikiness thought proper by later period groups and subsequently aped by “normal” symphony orchestras. To someone whose listening began in the 1960s, this is all to the good and this performance impresses by its sheer “rightness”. I am not sure what the historical justification for the fashionable spikiness is supposed to be, but for those who have grown up with it, what sounds like “rightness” to me might sound like “wrongness”.

One last point regards the minuet. Following with the Robbins Landon score, I found various extra ornaments in this performance – chiefly mordents and appoggiaturas. These are not Newstone’s embellishments: they are all in the old Bärenreiter edition. Evidently they are not in the earliest sources consulted by Robbins Landon, but since they are the sort of thing people did in those days, Robbins Landon might have left them in as optionals, at least in the repeats.

No. 34 was one of the symphonies completed by Max Goberman (1911-1962) before premature death cut short his project to record the entire cycle with the Vienna State Opera Orchestra. His tempo for the first movement is extraordinarily slow, making the most of the harmonic suspensions and passing dissonances and opening up a world of unexpected mystery. So far, so wonderfully good. Unfortunately, the harmonic scheme underlying the more florid second subject is much plainer. The result is that the violins seem to crawl through their semiquavers over a static bass line. What approached sublimity under Newstone becomes interminable. Goberman was known as a stickler for repeats, but neither is heard here – they would have stretched the movement to beyond seventeen-and-a-half minutes[3].

The second movement, with both repeats, is similar in tempo and playing style to Newstone’s, while the minuet has an affable gait with a particularly good trio – the syncopated horns well to the fore. The finale, with all repeats, is again similar in tempo and manner to Newstone’s. Were it not for Goberman’s ultra-slow first movement, I could be very happy with either.

Leslie Jones (1905-1982) and the Little Orchestra of London held an honoured place in the Haydn world in the 1960s. A pioneer of small band Haydn (and his contemporaries), he recorded almost half the symphonies for Pye and Nonesuch, but from the 1970s he and his orchestra quietly went off the radar and I have not been able to establish how long they continued to operate.

In the first movement of no. 34, he shaves three-and-a-half minutes off Goberman’s timing. This is definitely a three-in-a-bar performance which, with restrained phrasing and obvious deep feeling, avoids superficiality. He does not play either repeat and, unlike Goberman’s drawn-out performance, actually left me wanting more. That said, I do think that Newstone, at a tempo about half-way between the two, reveals an extra depth and, as I said above, sublimity.

Jones is unhurried in the second movement (and gives only the first repeat), but with punchy accents the music has plenty of life. Certain aspects of the playing look to the future – staccatos are a bit shorter and sharper, if not yet spiky. The minuet is rather formal, though alert, and the lively finale, like Newstone’s, has all repeats except the last. In the final resort, this is good rather than distinctive, but back in its day, when the choice was between this and Goberman, my choice would have been Jones. So on to the two cycles.

Märzendorfer takes the first movement fractionally slower than Newstone (but nowhere near as slow as Goberman), but to quite different effect. By making the most of dynamic contrasts and strong accents, he finds a note of tragedy in the music. Neither repeat is taken. His second movement, with both repeats, is not particularly fast but with strong accents, short staccatos and tangy oboes it emerges as terse and sinewy. The minuet struts pleasingly with a delightful trio. The finale, with all repeats, spins at a real Presto assai. If the other performances so far suggest a gigue tempo, this is a tarantella. Märzendorfer makes much of the dynamic contrasts, with real pianissimo playing where needed. A splendid performance all round.

If it troubles you that none of the conductors so far has played either repeat in the first movement, you will at least get the first from Dorati. It is a performance in the tragic mould, at a tempo similar to Märzendorfer’s, and he demonstrates that one repeat, at any rate, will not outstay its welcome. It is a more inflected performance, however, with tiny expressive hesitations emphasizing the harmonic shifts on the first page. The other conductors did not find this necessary, but I do not think Dorati overdoes it. His second movement is in a similar vein to Märzendorfer’s, though a little less tangy and with only the first repeat. The minuet has a delightful lilt – not to be taken for granted with Dorati – but the trio is a little plain. The finale goes at a high speed, but with tight control and even (and strong) accents on both beats of the bar in the lower strings, it is somewhat joyless. I didn’t think of the tarantella as I did with Märzendorfer.

A win for Märzendorfer, then, though I am happy to have the Newstone.

© Christopher Howell 2026

[1] Information from Gray’s Classical Discography https://classical-discography.org/  and WERM.

[2] See https://www.theguardian.com/news/2006/may/11/guardianobituaries.artsobituaries and, rather more detailed,  https://www.independent.co.uk/news/obituaries/harry-newstone-5336119.html, both retrieved 16 January 2026

[3] Not only would have, but almost certainly did. We have Anthony Hodgson’s assurance – The Haydn Seekers, CRC Magazine, Winter 2001 – that Goberman recorded every repeat in every symphony. These were then snipped off by the engineers to fit each symphony on an LP side. In principle, if the original tapes still exist, the repeats should be reinstated.