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. 64 in D major, “Tempora mutantur” Hob. I/64 (c.1773)

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

Though Wikipedia tells you the nickname is Haydn’s own, others are less sure. It appears on authentic manuscript parts in a Frankfurt library[1], but does this make it Haydn’s explanation of the symphony or somebody else’s comment on it? The source of the title was traced by Jonathan Foster to an epigram by John Owen (c.1565-1622), published in 1615:  Tempora mutantur, nos et mutamur in illis. Quomodo? Fit semper tempore peior homo (Times change, and we change in them. In what way? Time ever makes man worse). You may wonder how an epigram in Latin by a 17^th century Welshman came to be known in the Esterhazy court, but it seems this book was highly popular in European literary circles in the mid-18^th century.

Whether or not Haydn intended to express the idea that “times are out of joint”, or whether that was the impression one of his contemporaries got, it is not an implausible description.  The first movement begins gently and melodically. We might expect it to unfold lyrically in this manner, rather as did the opening movement of no. 29. Instead, after two bars it is rudely interrupted by an energetic forte passage. The gentle music resumes, is interrupted again and thereafter the energetic motive leads the proceedings. There is no real second subject. Just as one seems to be announced, the music enters what can only be described as a harmonic crisis. This is still disturbing today – it must have been devastating in 1770s Esterhazy. A few more brusque references to the energetic motif round off the first part. After a conversational development of the themes heard so far, a recapitulation seems to begin in the wrong key of D major. The actual recapitulation enters by stealth, halfway through a long paragraph.

The second movement is a broad and heartfelt Largo, initially played on just muted strings. A second movement for strings only would have been normal enough, but a little more than halfway through, the wind enter. There are no repeats – the music develops spontaneously, looking ahead to Brahms. All this would be remarkable enough, but the really curious thing is that the music keeps stopping. Not at logical phrases but in mid-phrase, almost as if it were accompanying a game of musical chairs, though young participants in that innocent party game would be puzzled by the serious tone of the music. Another remarkable feature occurs at the very end. The second horn has a pedal note at the bottom of its register, while the first horn rises above the strings, calling in the romantic age to come. Play this blindfold to a knowledgeable listener – but not quite knowledgeable enough to know this symphony – and they would be unlikely to guess at a date before 1850.

The minuet is notable for “Scotch snaps”, trills and grasshopper leaps in the melodic line, while the trio allows the high horns to shine. The finale is a rondo, but almost a parody of a rondo. The episodes try to develop a line of their own, but instead either stop in their tracks of peter out, lending a “back-to-the-drawing-board” feeling to each return of the main theme. All this creates an impression that things are not as they should be – an impression surely much stronger in an age that laid such store by formality. The first period of the rondo theme is also unusual in that it consists of two six-bar phrases.

The first recording was made in June 1950 by the Vienna Symphony orchestra under Henry Swoboda (Westminster WL 50-23 in the USA, Nixa WLP 5023 in the UK)[2]. The tempo for the first movement is cheery rather than urgent, and Swoboda obtains long lines and a singing tone from the VSO strings that is not always apparent in these early Haydn recordings – maybe because some conductors, even in the early 1950s, felt this was inappropriate to the classical style. Maybe it was, but it is a pleasure to hear, as is the use of high horns, though the players sometimes seem to be only just coping. Swoboda gives the first repeat and cuts two bars from the exposition – this may be down to the edition used rather than his own personal snip. The second movement is expressed broadly, rather romantically. In the minuet, the dance cadence, arising from the mix of Scotch snaps, trills and staccato notes, does not quite come off, but by the time they get to the da capo, the players have got the spring of it. Likewise in the trio, the high horns sound none too happy with their phrase the first time. This phrase, taking into account the repeats, comes four times, and gets better each time, till the last is very good. Studio time must have been at a premium, since everything would have been set to play it through again and give a really good performance. This is what rehearsals are for, but whether the repeats in a classical symphony should be used as rehearsals on a disc is debatable. The finale is unhurried but has plenty of high spirits and Swoboda seems to have understood the joke behind it. This performance gave listeners a good, if hardly definitive, idea of the symphony. Back in its day, a critic felt that “The conductor seems here to be trying to attain what his orchestra cannot or will not”[3].  Nevertheless, nothing else turned up till the middle of the next decade.

This was by the Little Orchestra of London under Leslie Jones (Pye GSGC-14045 in the UK, Nonesuch H-71121 in the USA), set down in 1965. The bright sound of Jones’s small orchestra is a pleasure to hear, and high horn parts evidently held no terrors for this specialist group. The first movement is taken briskly with fine appreciation of the dynamic contrasts. Neither repeat is given and there is the same two-bar cut as in Swoboda. The Robbins Landon Urtext came out that same year, but was evidently not yet available to Jones. The second movement avoids overt romanticism, but achieves a touching, heartfelt gravity. The minuet and trio go with real spring. The finale has good drive, though more might have been made of the obsessive oddness of this rondo – here perhaps Swoboda still has something to say. I have not always raved about Jones’s Haydn recordings, but this, in spite of the missing first movement repeats, is excellent, maybe ideal in the middle movements.  

The opening of the first movement, as played by Märzendorfer, is not songful but has the tension of a coiled spring. The ensuing forte truly explodes. The whole movement is tensile, breathlessly dramatic. Only the first repeat is taken – what a pity. However, he is using an Urtext edition, so no two-bar snip in the exposition. The second movement is made to sound, not an interrupted song, but is invested with a groping, uncertain air in which the pauses play their part. The entry of the wind is pure magic – it seems we are entering the Elysian fields, but then how shocking are the forte outbreaks. In Märzendorfer’s hands, this sounds to be one of the most profoundly unsettling slow movements in any symphony. The minuet is sprightly, with the high horns sounding lovely in the trio. The finale has a fiery drive, while the obsessive returns to the rondo theme are pointed in the way I missed with Jones. This is truly magnificent, even great in the second movement. I am afraid it completely eclipses memories of the Swoboda and Jones recordings, for all their merits.  

Dorati’s first movement is excellent, well phrased with plenty of vitality and forward movement. What I miss is the feeling that Haydn is straining within the constraints of the 18^th century aesthetic. The composer’s exposed nerve-ends, which Märzendorfer charts so well, are treated more politely. The first repeat only is played and, with Robbins Landon to advise, there is no two-bar snip. In the second movement, where Märzendorfer expressed a disturbing, uncertain world, with Dorati all is serene and, except where Haydn really insists with a couple of fortissimo explosions towards the end, is presented as one long, untroubled, even religioso, meditation. Can the music take this? Timings are dangerous tools but in this case they show that Märzendorfer, while giving the impression of feeling his way forward, is not really very different from Swoboda or Jones in terms of actual tempo, while Dorati is substantially slower than any of them. This is not Mahler or even Brahms, and the borderline between serenity and inertia is sometimes crossed, for all Dorati’s detailed shading. His minuet is rather formal, as is his manner, but this actually lets the oddities of the melodic line emerge with a droll humour. Why did I not enjoy this more? The answer seems to lie in the first note of the principal phrase. Dorati has quite correctly noted that this is not marked staccato, as are the upbeats in the following bars, and gives it its full value. Not a tenuto with a Viennese upbeat, just plonked there as is. The resulting lack of lift wraps itself round the theme like a wet lettuce, detracting from all that is good about the rest. The finale is like the first movement, well phrased and vital within 18^th century bounds, avoiding any suggestion that the composer is yearning to break loose. The really odd thing, though, happens at bar 125. Here, the music goes into the minor key, or at least, it does in the Robbins Landon edition and, to judge from the Swoboda and Jones performances, in earlier editions too. Under Dorati, it remains in the major, then flips into the minor four bars later. The result sounds weird, even for Haydn. However, since Robbins Landon was acting as editorial advisor for the cycle, we must suppose this reading derives from an authentic source. In which case, why does it not appear in his edition, even as a variant for conductors who prefer it?

In spite of a few reservations, the Dorati has a lot in its favour and might be favoured by those who do not wish their Haydn to be too challenging, though in that case I would press the claims of Jones. For me, Märzendorfer has so far exceeded his usually high standard as to silence all competition.

Or are there any live versions to take into account?

The Cameo Classics/Itter Collection set allows us to hear a performance that opened a concert by the BBC Symphony Orchestra under Hans Schmidt-Isserstedt on 20 November 1952. The other works on the programme were Egk’s French Suite and Hartmann’s Symphony no. 3. At this date we obviously have an old edition with the two-bar cut in the first movement. The first movement is taken at a fairly steady tempo, but with good care over phrasing and dynamics. The exchanges in the first part of the development are enunciated very clearly. The false recapitulation in the wrong key is heralded by a rallentando and a slight pause – not a bad idea considering that Haydn wanted to gull us into thinking this really was the recapitulation. At this tempo, there is a somewhat Mozartian feel to the music. The first repeat is played. The second movement is given with a heartfelt, long-drawn warmth that does not shy away from the strangeness of the pauses. The tempo is slow, but slightly less so than with Dorati – enough, it would seem, for him to get away with it. The minuet is surprisingly sprightly for this date. The high horns in the trio seem rather backwardly balanced, though they play beautifully. This could be down to the recording, but perhaps Schmidt-Isserstedt wanted this, since he brings out the concurrent detail in the string parts. The finale is good-humoured rather than dashing, but the conductor points well the obsessive returns of the rondo theme and his halting presentation of the final return is genuinely imaginative. If not challenging like Märzendorfer’s, this is a very nice performance, old-school Haydn at its best.

As I have remarked in other articles, SWR has issued a raft of Haydn symphonies with the South-West German Radio Orchestra under Ernest Bour, without it apparently occurring to them that this sort of historical issue appeals particularly to collectors who like to have information about dates and venues. Bour’s recording of Symphony 64 may date from 12 March 1960[4]. Inevitably, the old score is used with the two-bar snip in the first movement. This latter is brisk and bright, with phrasing and dynamics well observed, though with no special features, except the omission of both repeats. The second movement, too, is attentively played while leaving me with the feeling I had heard better elsewhere. When the wind enter, the oboes are so backwardly balanced that the magic is lost. Things buck up in the bouncy scherzo, with the high horns bold rather than lyrical in the trio. The finale has plenty of drive. If issued back in 1960, this would have been most welcome, and even in 1965 would have been a valid alternative to Jones for those who prefer a slightly larger orchestra, the more so when Jones also omits both repeats in the first movement. It was comprehensively superseded by Märzendorfer.

* On its first statement, the rondo theme has two brief repeats. All these performances play them.

Christopher Howell © 2026

[1] H. C. Robbins Landon: Haydn. Chronicle and Works. Vol.II. Haydn at Esterhaza  1766-1790 Thames and Hudson 1978, pp. 285 and 307.

[2] My article on Swoboda was the first in my series of “Forgotten Artists” and, even with its subsequent updating needs further revision in the light of information and further material that has come to light since. However, it remains a fair introduction to this conductor. 

[3] C. G. Burke: The Music of Jos. Haydn on microgroove records, High Fidelity November-December 1952, p.73.

[4] A recording by this orchestra and conductor and with this date was once available from René Gagnaux’s site “Mon Musée Musical”. Unfortunately this site has experienced problems and at present (16 May 2026) the links do not work, so I cannot say if this is the same performance.