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. 73 in D major Hob. I/73 “La Chasse” (1782)

Though “La Chasse” was not Haydn’s own name for the symphony, it has more justification than most such titles since the last movement, originally used as the overture to the opera La fedeltà premiata, clearly contains hunting horns and actually quotes from La Chasse du cerf by Jean-Baptiste Morin. Most unusually for a Haydn finale, it finishes piano. You would be hard put, however, to find a hunting programme in the rest. The first movement has a slow introduction that begins with slowly pulsing repeated chords in the wind, against which the violins outline fragmentary motifs. After climaxing in expectant manner on the dominant, the main Allegro dashes off, again with repeated notes, this time in stuttering quavers, without really establishing the tonic key. There is no second subject proper, the repeated notes dominate all and everything in this first part leads to a conclusion in the dominant. The development goes through some surprising modulations before repeating the opening gambit of climaxing on a dominant chord, after which the recapitulation slips in off-key as before. So the tonic is only established firmly at the very end. I apologise for the technical details, but even those who do not understand them may grasp intuitively that this apparently delightful, playful movement is actually rather nervous and unsettling. The second movement, which draws upon an earlier song by Haydn, alternates gentle grace with, in the minor key episodes, a sadder elegance. The minuet has an Allegretto marking and, while it is clearly not a one-in-the-bar scherzo, it seems to benefit when not taken too slowly. The finale lives up to its name and is, apart from the delicate ending, a jolly romp. This is compounded by a Haydnesque surprise that disappeared from view until the Robbins Landon and other Urtext editions came along. Having drafted it in from an opera, Haydn left the trumpets and drums in the orchestra even though it meant they sit through the first three movements playing nothing. Quite a shock when they come in, especially if the conductor gives them their head. The logistics of this evidently worried earlier editors, who removed them from the score. So most of the conductors discussed here were unaware of the trumpet and drum parts.

The earliest complete traversal (Hans von Benda had recorded the finale in the 30s) was set down on 8 February 1945 by Fabien Sevitzky (son of Serge Koussevitzky) with the Indianapolis Symphony Orchestra. This is available from Pristine Classics, in a transfer by the experienced Mark Obert-Thorn, as part of their comprehensive survey of the Sevitzky/ISO legacy. The sound is rather in-your-face in a dry acoustic, but I imagine Obert-Thorn has done his best with the material. Sevitzky moves the introduction on, emphasizing the quirkiness of some of the phrases. In the Allegro he goes for robust rather than vivacious energy, but I found this effective. Neither repeat is played. The Andante seems rather heavy-handed, though the real issue may be that his players are not very good at suggesting long cantabile phrases even when the notes are staccato. They come out as a series of single jabs. He does show appreciation of the more dramatic outbursts later on. No repeat. Sevitzky’s minuet is a very brisk Allegretto indeed, with the trio sounding perfunctory. It may not be just a matter of tempo, if only it were done with a touch of elegance. There are Haydn minuets that work as stomping peasant dances, but this one does not seem to. The finale is a lusty, even rowdy tally-ho affair, but I do not get much sense of enjoyment, more the impression of a conductor goading them on with an iron grip. For all its energetic merits, I doubt if those who bought this because it was the only recording thought the last word had been said.

In three movements at least, Harry Newstone’s 1952 recording with the Haydn Orchestra (HSLP 1052/Nixa NLP 902) was a considerable improvement[1]. Newstone plays the first movement introduction broadly, followed by a thoroughly vivacious Allegro with some brilliant playing. He plays the first repeat. In the second movement he captures beautifully both the gentle charm of the major key sections and the sad, yet poised, grace of the minor key episodes. He ignores the single repeat. Where I have doubts is in the minuet. Unlike the corresponding movement in no. 49 on the other side of the disc, you would not doubt from his performance that it is a minuet, but modern ears will find it very slow, for all its delicate poise. Not so the “hunting” finale, which spins along with much verve. The single repeat is played.

The 1950s were not prolific in further recordings. I have traced only one, set down on 5-10 July 1957 by Rudolf Alberth with the Lamoureux Orchestra for the Club du Disque Français (RC-869). Albert(h) (1918-1992) has something of a niche reputation. Born in Frankfurt he had a major career in Germany but was also active in Italy and France. He seems to have added the “h” to his surname for the benefit of his French audiences, who would otherwise not have pronounced the final “t”. He conducted a wide repertoire with a particular interest in contemporary music – he gave the première of Messiaen’s Oiseaux exotiques. He plays the introduction to the symphony fairly swiftly, with a fine sense of expectancy. In the first movement, he manages to combine Newstone’s vivacious energy with Sevitzky’s robustness. He charts well the surprises in the development. This sounds ideal to me. The first repeat is played. The Andante is faster than the previous versions, reasonably so since it is in 2/4 not 4/8. It is delightful, yet finds room for the stronger moments. This sounds ideal, too. He plays the repeat. His minuet is a vigorous, one-in-a-bar scherzo, even suggesting anticipations of Beethoven 9 in the hammering octaves. It works splendidly, but can it be right? As if in answer to my question, he takes a slower tempo for the trio. Allowing the legitimacy of this, he does it with a piquant charm. The finale, with repeat, has dash and excitement, without ever losing control. This is splendid, but once again the minuet may be the issue.

The symphony fared a little better in the 1960s. First came a version issued by Bertelsmann (14 383, also issued in the UK on World Record Club ST 206) and played by the Westdeutsches Mozartorchester conducted by Walter Schulten. No other recordings by this orchestra or conductor seem to exist and it is difficult not to suspect a pseudonym, although Bertelsmann generally engaged genuine if obscure artists and the other side of the disc had a performance of Symphony 45 by the Südwestdeutsche Kammerorchester under Friedrich Tilegant, a pupil of Hindemith. A Walter Schulten was prominent in the Bruckner world for his investigations into the composer’s apprentice years[2], so if this was the same person, perhaps he was a musicologist with some ambitions to be a conductor. There is certainly a musicological rigour to the performance that separates it from others I have heard. The repeated chords that open the work are spelt out six-in-a-bar and rather heavily, as though the conductor wishes us to understand that repeated chords in some form or another dominate the entire work. The following Allegro is taken fairly slowly but this is where things become interesting, for the repeated quavers are hammered out with great vitality while there is a fiery articulation to the semiquavers. Schulten finds real strength in this movement. The first repeat is played. The second movement is taken definitely four-in-a-bar, and my immediate reaction was that this was going to be too slow and heavy. Certainly, there is little charm or grace, but here too, the conductor’s seriousness becomes almost riveting. The repeat is not played. The minuet is quite fast, not perhaps a one-in-a-bar Beethoven scherzo, but plausibly anticipating Bruckner. The trio does sound heavy and charmless, though. The finale avoids the bouncy and élan of a hunt to concentrate on musical values. At a steady tempo, and with the repeat, the single quavers are again played with such energy and fire that the result is ultimately convincing. Indeed, such is the power and authority of this distinctly austere performance, which demonstrates that Haydn can be conducted like Horenstein conducting Bruckner and emerge remarkably well from the experience, that I am left wondering if it is actually the work of somebody quite important who could not be named for contractual reasons.   

The conductorless Prague Chamber Orchestra set down their version for Supraphon on 14-15 December 1961 as part of a series that paired lesser and better known works – no. 73 was coupled with no. 96. The introduction is taken slowly and the pulsing chords are separated more than in any previous performance. The result is rather static. The Allegro goes at a lively pace and here we begin to note the difference between playing with and without a conductor. The string playing is brilliant and there is a feeling that the violins are leading the way, pacing the music in terms of their instrument rather than adapting to somebody else’s idea. With plenty of vitality and both repeats, this is all to the good, though it could be felt that an experienced conductor such as Alberth times the stoppings and startings in the development more communicatively. The Andante shows another aspect of playing without a conductor. When the melodic interest passes from one section to another, you get the feeling of each section taking the initiative in its turn. The effect is more that of an enlarged string quartet than of a reduced orchestra. This, too, is all to the good. They take this movement, with the repeat, almost as slowly as Newstone and Sevitzky, definitely four-in-a-bar, and after Alberth I tend to miss the swifter tempo, for all the elegance and poise of the Prague players. The minuet is slow, inflected with many airs and graces – too slow for me, though I did enjoy the Trio in which the Czech wind players display their rustic charms. The really odd thing is that, having taking all repeats up till now (and they take the repeat in the finale), they omit the second repeat in both the minuet and the trio. I suppose the engineers might have chopped these off to fit the performance on an LP side, but surely LP sides could stretch a couple of minutes beyond 25 by 1961? The finale is lively but, without exactly slowing down, there is a sensation of losing steam here and there in the quieter passages. This is perhaps the downside of having no conductor. It is the difference, perhaps, between a horse galloping along for reasons of its own, in its own good time, and a horse goaded ever on by the merciless rider on its back, or, in the case of the orchestra, by the sod on the rostrum. For all its charms, this pales a little beside earlier versions.

In 1964, the Canadian company Baroque Records issued a pairing of symphonies 70 and 73 by the Dresdener Kammersolisten under Marcel Bernard (BU 1823). The conductor’s name arouses some suspicion since Music for Pleasure, in its earliest days, issued a series of recordings in which a conductor of this name directed the “Westminster Symphony Orchestra” in repertoire that looks like pseudonymous borrowings from the Everest catalogue – the slightly unusual pairing of Mozart 40 and Schubert 8, for example, mirrors a coupling by the LSO under Leopold Ludwig for that label. I have not been able to check out any of these MfP releases, so my doubts remain doubts. In the case of Haydn’s 70^th and 73^rd, it is difficult to see what they could have filched at that date. The Canadian company issued a few more records by this combination, and the component parts show some sign of independent life. The orchestra made an LP or two with other conductors and Bernard also recorded with the Baroque Chamber Orchestra of Hamburg, a genuine ensemble founded by the trumpeter Adolph Scherbaum. So I will assume he was real and the misuse of his name by MfP a pure coincidence. This would matter more if the performance was better. In truth, it sounds rather provincial, though Dresden is not, culturally or musically, a provincial city. The chording in the introduction is imprecise and the playing directionless. The Allegro, with no repeat, goes at a pretty comfortable tempo, but Haydn’s good humour takes care of that and the result is enjoyable in a modest way. Likewise the Andante, nicely turned at a good mid-way tempo. The minuet is not one of the slowest and the tempo might have done, but there is no lift to the playing, which is leaden in the extreme. The finale, without repeat, goes with considerable verve but then, I do not think I have heard any version of this symphony where the finale did not come off, whatever happened elsewhere.

The series of Haydn symphonies made by David Blum and the Esterhazy Orchestra for Vanguard were well regarded in their day. They set down no. 73 on 24 October 1963 and 6-7 February 1964. His slowly pulsing wind at the beginning are similar to Newstone’s. The Allegro proper is fractionally slower than most, concentrating on friendly good spirits rather than vivacity, though the strings are as brilliant as any. Blum has the wind well to the fore and their colourings give the music a pastoral air not heard so far in this listening. Only the first repeat is played. The Andante is a bright, alert two-in-a-bar but, much as I favour this in principle, unlike Alberth’s performance, you get the impression that he is keeping it going at all costs. It lacks repose. The repeat is played. The minuet is a euphoric one-in-the bar scherzo. Unlike Alberth, he keeps his tempo for the trio, dismissing it all too hastily. Alberth’s slowing and Blum’s haste seem to prove that, while the minuet might work as a scherzo, the trio does not. The finale, with the repeat, is, like the first movement, sufficiently steady to suggest boisterous good humour rather than a dash for the finishing line. After a particularly good first movement, this performance lives up to its promise only in the finale.

At an unspecified date in 1964, Leslie Jones recorded his version with the Little Orchestra of London. His introduction is rather pedantically spelt out but the Allegro spins along at a vivacious lick. With a lack of real phrasing, it sounds breathless at times. His middle movements are slow and very slow, without the delicate grace with which Newstone could vindicate such tempi in the Andante and, up to a point, in the minuet. An enjoyably lively finale does not really save the day. First repeat in the first movement, no repeat in the second, repeat in the finale.

The Hungarian Chamber orchestra, led by Vilmos Tátrai, set down a number of Haydn symphonies in the 1960s and early 70s, cunningly mixing better-known and rarer works, mostly from Haydn’s middle period. Tátrai formed the Tátrai String Quartet in 1946 and the Hungarian Chamber Orchestra in 1957. In CD reissues of these recordings, Tátrai is usually said to have “directed” the performances, implying that he conducted them, but the original LPs more clearly described him as the leader – this orchestra played without a conductor. The pulsating chords at the beginning of the introduction float by very serenely and even the few forte moments have a soft attack that does not disturb the mood of intimate contemplation. The Allegro then scampers along with a sense of the players delighting in their own virtuosity – not for the first time, a conductorless chamber orchestra sounds like an enlarged quartet. The first repeat is taken – I was ready to enjoy hearing the second part over again. My only other complaint is that the wind are rather backwardly balanced. So far, so delightful. The Andante is played so fast as to have, for me, something of an Allegretto feeling. Some will find it absolutely charming, but too many performances have unearthed depths beneath the apparently simple writing for me to find this wholly satisfying. The repeat is played. The minuet is nicely done but lacks the spirit of the dance and the finale, given rather steadily with the repeat, lacks the spirit of the hunt. I find it puzzling that the players could show such verve in the first movement and then seemingly lose interest in the music.

Like Alberth, Leopold Ludwig (1908-1979) had a niche reputation. He made a few records with the Philharmonia Orchestra in the 1950s, but Walter Legge evidently thought him no more than a safe pair of hands when there was a bit of accompanying to be done.  His coupling of Haydn’s 31^st and 73^rd Symphonies was set down on 6-7 April 1966 with the Bavarian Radio Symphony Orchestra for EMI Electrola and does not seem to have been issued outside Germany (it is not listed in Gray’s Catalogue) or to have achieved a CD transfer. The introduction pulsates nicely, with proper appreciation of the quirkier phrases. The Allegro goes at an ideal tempo that simply beams enjoyment in the playing. The first repeat is played – the second would have been more than welcome in such a performance. The Andante is just fast enough to suggest two-in-a-bar but with all the time in the world – a sort of pre-Schubertian amble. Ludwig seems to be using a fairly full orchestra, but is able to have them playing lightly with clear textures. The minuet is only a little faster than the slow minuets of Newstone and others, but oh what a difference that little makes! In place of powder-puff-bewigged curtseying (or a proto-Beethovenian scherzo), we have a one-in-the bar, lilting Landler. The trio is enchanting. The finale struck me as a tad slow for a Presto, but the playing exudes enjoyment and I must say that, at this tempo, the acciaccaturas in the hunting horn motif sound like genuine grace notes and not mistakes. A lovely performance.

Turning now to Märzendorfer, the first question is the edition used. All performances discussed so far depart from the Robbins Landon edition in various details of phrasing and dynamics, and since they all depart from it in exactly the same way, I take it they are all using faithfully the only edition available till the Robbins Landon hit the dust in 1967. These are not drastic differences, though it would be better to comply with them now the Urtext is readily available. The drastic difference is that trumpets and timpani, hitherto silent, turn up in the last movement of the Robbins Landon edition, and if they played in any of the recordings I’ve heard so far they were very quiet about it. With Märzendorfer, they let fly with all the enthusiasm of players who have been sitting there doing nothing for three movements. IMSLP has an undated edition published in Leipzig that has no trumpets and drums, and I imagine this is the edition the earlier performances used. However, Märzendorfer differs from Robbins Landon on two points. At the very beginning of the symphony, Robbins Landon has the lower strings marked pizzicato, but not the violins. This is how it is traditionally done. Märzendorfer, or his editor, takes the pizzicato as applying to all the strings. This makes a big difference. Playing with a bow, the violins sketch out a theme against the pulsing wind chords. Playing pizzicato, they are barely audible. Since the chords in themselves do not form a theme, the effect is very odd. The other point is the beginning of the minuet, and every time the initial theme recurs. Robbins Landon has forte in square brackets, indicating that this is an editorial addition and that Haydn put no marking before the sudden sforzatos in bars three and four. The old Leipzig edition had a forte marking. Authentic or not, it seems logical, and every other conductor has followed this. Märzendorfer begins piano, bursting abruptly to forte in bars three and four. It sounds quirky, but quirkiness is nothing new in Haydn.

Apart from the oddity of the pizzicato violins, Märzendorfer plays the introduction in a way that avoids too much obvious continuity, emphasizing the stopping and starting and the odd chromatic touches. His Allegro, with rustic winds well to the fore, comes as a joyous release, but he also finds mystery in the development. Both repeats are played. His Andante goes at about the same tempo as Ludwig’s, but with sharper staccatos he avoids straying into the world of Schubert. The repeat is played. His minuet, too, goes at a tempo not far different from Ludwig’s, but with the emphasis on quirkiness and sudden contrasts. No relaxed Ländler here. The finale, with ringing trumpets and hard-sticks on the timpani (and the repeat) is a rumbustious affair like none other so far. I shall keep the Ludwig by me for more relaxed enjoyment, but for a demonstration of Haydn breaking out of conventional 18^th century moulds, Märzendorfer is (so far), hors concours.

Since Dorati’s cycle was declaredly advised by Robbins Landon, it is not surprising to find him complying with that edition on the points above. Incidentally, one other point in the Robbins Landon edition, followed by both Märzendorfer and Dorati, is that the horns and oboes have appoggiaturas not acciaccaturas, which is easier to handle without them sounding like mistakes. Dorati is less radical than Märzendorfer. He gives the introduction more continuity and the Allegro, with both repeats and less forward wind, bowls along in a sprightly manner. The Andante is a perky strut, in an 18^th century garden rather than the countryside. The minuet, slower than Märzendorfer (but not as slow as some others), has slightly affected phrasing creating an exaggeratedly rococo effect. I was glad when it was over. The trumpets and drums are kept at bay by Dorati, but it cannot be said they are inaudible and this infectious account shows the Dorati cycle at its best. A pity about the minuet. 

Just one live version caught my attention – Paul Paray included the symphony in a concert with the Detroit Symphony Orchestra on 26 November 1959. Paray was famous for his fast tempi and he plays the introduction pretty swiftly, though this does not prevent him from pointing the quirky chromaticisms sensitively. Likewise his Allegro, with just the first repeat, spins along at considerable speed, but does not sound breathless and the surprises in the development are properly registered. The Andante, without repeat, is two-in-a-bar but unhurried, a perky strut not unlike Dorati’s. His minuet seems to me exactly right – one in a bar but not a pre-Beethoven scherzo. It has an infectious bounce. The finale, with repeat, is unexpectedly slow, but it manages a punchy zest nevertheless. The pre-Urtext edition is obviously used. A pity Paray was not asked to record some rare Haydn during his Detroit years.

The big issue here is the trumpets and drums in the last movement. If you want those, then within my cut-off date Märzendorfer stands alone. Leaving this aside, Ludwig offers a lovely old-style performance while I will certainly want to hear Alberth and Paray again.

© Christopher Howell 2026

[1] See my article on Symphony 34 for a brief profile of this conductor

[2] For example, Anton Bruckners künstlerische Entwicklung in der St. Florianer Zeit, 1845-1855 (Dissertation, Mainz University 1956)