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. 13 in D major Hob. I/13 (1763)

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

With flute and drums in addition to the usual oboes, horns and strings, Symphony no. 13 might be classed as a festive symphony. It is also, if to a lesser extent than the trilogy formed by Symphonies 6-8 (“Le Matin”, “Le Midi” and “Le Soir”), a concertante symphony. The Adagio cantabile second movement consists entirely of a florid cello solo, accompanied very lightly by just the other strings. The flute, which has not had much to do so far, leads the trio of the third movement with an agile solo spot. The timpanist, too, has a moment of glory with his four quavers at the end of the first part. This symphony has sometimes been called the “Jupiter” since the four-note theme in the finale is identical to that of Mozart’s “Jupiter”. It is not developed fugally, however, whereas the similar but not identical four-note theme in the finale of Symphony no. 3 combines fugal and sonata form in a manner that strongly anticipates Mozart’s finale. Logically, then, if you must attach the nickname “Jupiter” to one of Haydn’s symphonies, it had better be the third. Nonetheless, though the finale is not fugal, it is more contrapuntal than it may seem at first. This is soon evident when the theme passes to the lower strings and the upper strings take up the quaver motif that might, at the outset when played by the lower strings, have seemed pure accompaniment, but now turns out to have been a counter-melody. Haydn makes much of this in the development and it is there that the music most nearly resembles Mozart’s finale in its actual sound.

The first two recordings must have been made at about the same time, but that by Jonathan Sternberg and the Vienna Symphony Orchestra (1949, HSLP 1001) came out first. Sternberg’s first movement struck me as more Allegro maestoso than Allegro molto, but the playing is punchy and energetic as well as suitably festive. In the Adagio cantabile, he has a cellist of considerable repute in post-war Vienna, Nikolas Hübner (1910-1985). Among other recordings, Hübner was one of the players in Clifford Curzon’s famous recording of Schubert’s “Trout” Quintet, as well as recording Mozart’s Piano Trios with Lili Kraus and Willi Boskovsky. It becomes evident, though, that one of the difficulties of this movement will be for the accompanying strings, which have solid quavers most of the time, not to weigh too heavily upon the soloist’s flights. This is not entirely avoided here, nor does Sternberg always keep soloist and orchestra together when Hübner allows himself a minimal touch of rubato. The minuet is not as heavy as some of Sternberg’s can be, but it is rather stolid and does not find the humour in the kangaroo leaps and Scotch snaps. The flautist is excellent in  the trio and it is a pity that, having named the cellist, the LP does not credit the flautist. It does, however, name the harpsichordist, none other than H. C. Robbins Landon, in spite of the fact that the recording does not register his presence so far as my ears can tell. The timpani sound a bit cavernous in their four notes at the end of the first part of the minuet, but I doubt if these will register properly until someone records them with hard sticks. The finale, like the first movement, is not quite Allegro molto, but has sufficient vigour and drive to come across well. A decent, if hardly definitive, first recording.

The performance by Hans Weisbach and the Bamberg SO (Mercury MG 10079) came out a year later but was likely made in 1949. It was part of a slew of recordings obtained from German and Austrian radio stations at the start of their operations, enabling them to build up a catalogue swiftly till they were ready to branch into products of their own, leading to the famous recordings made in Minneapolis, Detroit, Chicago and also London. I was appreciative rather than bowled over by Symphony 6 on the other side, but here I found more to enthuse about. Once you have become acclimatized to the fact that the recording itself is smoother but duller than the Sternberg, the initial impression of a lack of bite turns into one of exhilaration. I could have sworn that this was faster than Sternberg, but the stopwatch says there is only four seconds’ difference. Weisbach is less punchy, less military, less festive in a way but, with lighter accents, longer phrases and subtler phrasing, the result is delightful. His cellist, uncredited, in the Adagio cantabile is at least as good an artist as Hübner, though after the double bar he indulges in a burst of romantic passion that would be more suitable in Dvořák. Weisbach, though barely remembered today, was obviously a conductor of some skill and he keeps the simple accompaniment afloat, allowing the soloist space to express himself freely but, except in that one moment, tastefully. It is well that this is so, since Weisbach allows the first repeat, surely unusual at the time. His minuet is swifter than Sternberg’s. If not quite humorous, it is elegant and light on its feet. A fractionally slower tempo for the trio permits the flute to point the music with much charm. With a lighter touch in the finale, Weisbach does not labour the “Jupiter” connections. I am not sure that the tempo is much faster than Sternberg’s – this would require difficult calculations since Weisbach makes both repeats – but, as in the first movement, he is less punchy but more exhilarating. Decent as Sternberg was, this is much better.

I suppose some might feel that Weisbach is a little too Mozartian, in which case enter Max Goberman and the Vienna State Opera Orchestra. This was recorded between 1960 and the conductor’s early death at the end of 1962, when he was rather less than half way through his projected Haydn cycle. His first movement might be called improved Sternberg. It is majestic, but punchy and lively and better recorded with the forte-piano contrasts registering properly. The second movement is taken very slowly and tenderly. The cellist plays well but either does not have a real solo personality or is being overly careful not to make it sound like a romantic concerto. As a result, I became increasingly conscious of the length of the movement, and this without either of the repeats which Goberman almost certainly played. Goberman’s minuet is slightly pompous, but in a cheery sort of way – it is not heavy. The flute is very neat in the trio, but gives the impression of taking great care to get all the notes right – it does not sound very spontaneous. The finale is taken rather grandly, as if Goberman wishes us to hear the “Jupiter” connection, but it is lively and buoyant. Overall, then, improved Sternberg but maybe not improved Weisbach.

Leslie Jones and the Little Orchestra of London followed in 1965. With his little band, the wind are well forward in the first movement – the flute sounds delightful. The harpsichord is not so far forward but, unlike so many of these recordings, we do at least register that it is there. The playing is springy and vital. In the second movement, Jones takes heed that the accompanying quavers are marked staccato. This is all very well, but the result is that they seem to be actually working against the lyrical cello part, which is very nicely played. Another result is that, though Jones is almost a minute faster than Goberman, he sounds slow. Haydn wrote staccato, and staccato means staccato, but there would have to be a better way of doing it. In the minuet, the forward wind mean that some of the string lines, which carry the melodic interest, are obscured. The tempo seems to me on the fast side – perhaps the strings would have got their notes through better at a slightly slower tempo. The effect is jolly but breathless. Jones has to take the trio a little slower – this is delightfully done by the flute. As ever with slower trios, there is an uncomfortable jolt when the minuet comes back. The finale, like the first movement, is upfront and zippy – not much “Jupiter” grandeur here, but who says there should be? The harpsichord seems to be enjoying himself no end in these last two movements – not much of his work is clearly audible on the recording but it might be too much of a good thing if it was. The best outer movements so far, some reservations about the inner ones.

In 1967, Turnabout issued a recording (TV 341285) in which Antonio Pedrotti conducted the Bolzano Haydn Orchestra – an Anglicized and shortened name of the Orchestra Haydn di Bolzano e Trento/Haydn Orchester von Bozen und Trient (the Trentino-South Tyrol region of Italy has semi-autonomous status and is largely German speaking). Pedrotti (1901-1975) was a pupil of Respighi and has something of an underground reputation with collectors, deriving from his recordings for Supraphon, many of them with the Czech PO. He was born in Trento and remained attached to his region of origin. When the Orchestra Haydn was created in 1960, he was appointed its first principal conductor, holding the post till his death. This orchestra has a good reputation in Italy and has seen some distinguished visitors over the years, both on the podium and as concerto soloists. It has never enjoyed a high international profile, and has perhaps never sought one, since its declared mission was to bring music to its region. True to this vision, Symphony 13 had a string of performances in 1966, on 28, 29, 30 and 31 in Bressanone, Siusi, Salò and Riva del Garda, and, as part of a different programme, on 11 and 12 November in Trento and Bolzano[1]. I have not found the exact date of the recording, but it seems likely that it was made in the summer recess following the July performances. For all but the Salò outing, the solo cellist in the second movement is named as Umberto Ferriani. The soloist is not credited for the Salò performance or the recording, but we can surely presume that Ferriani played for these also. The recording shows a very high standard of playing and it is strange that Vox/Turnabout did not develop this association any further.

But first, a word about the edition used. Two features stand out at variance with previous recordings: there are no timpani, and the opening of the first movement is played piano. The edition is not declared but evidently the Robbins Landon score, published in 1963, was either not used, or adjusted by an editor who did not agree with all of Robbins Landon’s conclusions.

Regarding the timpani part, Robbins Landon tells us that it “was added [in the autograph] at the bottom of each page on a free stave; but not in Haydn’s hand. The ink suggests, however, that it was added early, possibly by the timpanist Adam Sturm, perhaps under Haydn’s supervision … we have evidence that the timpani part is contemporary, for we find it in several copies [he lists three]”[2]. It seems difficult to disagree with Robbins Landon’s evident conclusion that Haydn authorized and approved the timpani part, but I suppose there might be a certain musicological rectitude to the more Spartan argument that, since the symphony was seemingly conceived without it, we should not play it. The effect is slightly less festive, but fresher and more exuberant, at least as Pedrotti plays it.

As for the piano beginning, the Robbins Landon edition has forte in all the parts in square brackets, indicating that Haydn actually wrote nothing. The old Breitkopf edition has forte without any suggestion that this is editorial. Given that Haydn wrote nothing, would it not be logical to begin forte, especially when the orchestration is very full? Probably yes, but the justification for playing it piano comes at the moment of recapitulation, where the string parts are marked piano[3]. Could this justify retroactively placing a piano marking on the same music at the beginning? Robbins Landon was quite sure it did not – “The recapitulation brings in the whole first subject piano instead of forte”[4]. He might have added that the timpani, which underpin the first beat of each bar at the beginning, are silent at this point in the recapitulation, but this argument might not appeal to a musicologist who has rejected the timpani anyway. Though I would side with Robbins Landon over the forte beginning, this performance shows it can sound enchanting piano, especially when no timpani are present.

Leaving aside musicological matters, the performance is exceptionally fine. The first movement is a joyous ride through the open air, with the occasional flute interjections well to the fore. In the second movement, the cellist offers more detailed phrasing than any heard so far, yet without straying into romantic pastures. Pedrotti has the quaver chords separated, but as a delicate pulsation that supports, rather than weighs down, the soloist. His minuet is slowish and slightly pompous, but with astute timing he gives it an ironic air, a parody of a pompous minuet rather than real thing. In the trio, the flautist combines grace with a humorous smile. I would say that Pedrotti finds more in this movement than his predecessors on disc. The finale goes at a terrific pace, a real opera buffa finale. This is by a long way the best performance so far, so what do the two complete cycles bring?

Märzendorfer has the Urtext with timpani and a forte beginning. The timpanist is using hard sticks by the sound of it, and this first movement gets an exhilaratingly festive account. Both repeats, too. The cellist in the Adagio cantabile is trying almost too hard, with some uneven bulges to his tone. Märzendorfer has the string chords separated rather deliberately and the result, with the first repeat, is a little laboured. The minuet is slowish and almost military in its cut, lacking the humour of Pedrotti. The trio, too, is rather stiffly done. The finale, with both repeats, is notable for the rigorous clarity with which Märzendorfer brings out the four-note theme wherever it appears in the texture. The performance is punchy and energetic, but a little of Pedrotti’s verve would not have come amiss.  

Dorati’s cycle had Robbins Landon as editorial advisor, so no surprise if he has timpani with hard sticks and a forte beginning. He is a tad slower than Märzendorfer and is perhaps over-emphatic with his accents. I thought the way he has the horns blast out their arpeggios after the piano recapitulation rather vulgar, but some might find it good fun[5]. His account is certainly festive. One thing that might be said in favour of Dorati’s second movement is that it sounds like an Adagio in four, as written, not in eight. In fact, his performance with both repeats takes only two seconds longer than Märzendorfer’s with only the first. But the proof of the pudding is in the eating and the cellist does not really have time to express the music. His occasional attempts in that direction have to run the gauntlet of Dorati’s rigidly metronomic spelling out of the quaver chords. The result is humdrum. Dorati’s minuet is the slowest of all, a humourless and formal rendering. The trio is no better. The finale, on the other hand, is splendid, with a verve that comes close to rivalling Pedrotti. All the same, two good outer movements do not save the middle ones, or the performance as a whole. Textual matters apart, it is clear that nobody up to that time had matched Pedrotti for insight, character and verve in this symphony.

I have just one live performance to comment on – by the Boston SO under Erich Leinsdorf on 31 March 1967. Early Haydn does not easily connect with the image of Leinsdorf that posterity has handed down to us, but remember that he was the first conductor to record all the Mozart symphonies and a search in the BSO archives[6] shows that he conducted quite a lot of Haydn during his tenure and by no means stuck to the well known works. This was the first Boston SO performance of the symphony.

Leinsdorf conducted from the harpsichord, something that surprised critic Arch L. Crossley: “The program started a few minutes after 8:30 when Erich Leinsdorf came on stage, sat down at a harpsichord, and began to conduct Haydn’s Symphony no. 13 in D. You heard me: he sat down at a harpsichord. Seems that Papa Haydn used to do this, back in the early days at Esterhazy when this symphony was composed. He would strum chords with one hand and conduct with the other. There were probably some differences. For one thing it is doubtful that Haydn had 70-odd musicians with him, so that the harpsichord could be heard. Last night it was barely audible beyond the first two rows. Of course the symphony is usually conducted conventionally from the podium, but Mr. Leinsdorf is not above a little chaste showmanship.[7]” Another critic, Thomas Willis, praising the “sprightly performance”, also noted that “Mr. Leinsdorf conducted from a harpsichord, which demonstrated his stylistic awareness, altho he played very little”[8]. The recording, which can be heard in streaming upon request to the BSO archives, confirms that Leinsdorf either played very little, or could not be heard, or both. It also shows that, if Leinsdorf really had 70-odd players, he obtained the clarity and precision of a small chamber group.

The programme leaflet is silent over editions. This series of articles is revealing that few Haydn performances correspond exactly to any given score. This one seems to follow the Robbins Landon Urtext in the first two movements, after which a number of oddities appear. In bars 11-12 of the minuet, and in the corresponding bars in the second part, the strings play an unscripted pizzicato and in bars 39-40 the flute part is adjusted to match the first time round. In the finale, in the passage from bar 33 onward, Leinsdorf adds some clever echo dynamics and he apparently likes the last four bars of each section so much that he plays them twice over each time, with a Rossini-style crescendo. It is of course possible that these are not Leinsdorf’s own amendments – perhaps some musicologist has found a contemporary copy which has these features, but it would be nice to know.

That said, the performance is a delight. The first movement is not especially fast, but has a jaunty, upfront feeling. The Adagio cantabile is played with great artistry by Jules Eskin[9], finely supported by Leinsdorf’s gentle string chords. The soloist seems to have all the time to express the music, so it was a surprise to find that this is among the faster performances. The minuet, too, finds all the time to bring out the humour as only Pedrotti had among recorded performances. The flute is spritely in the trio. The finale is full of high spirits. Whatever the musicological niceties, the performances I shall return to are this one and Pedrotti’s.

The BBC took early notice of Symphony 13 – W. K. Stanton conducted a performance with the BBC Midland Orchestra in 1939. They also included it twice in the 1965 series featuring all of Haydn’s symphonies – by Harry Newstone and the BBC Welsh Orchestra and by Graham Treacher and the BBC Scottish Orchestra. A further performance was given in 1970 by John Carewe and the BBC Welsh Orchestra.

Christopher Howell © 2026

[1][1] Full details of the orchestra’s public concerts may be viewed at its website: https://www.haydn.it/orchestra

[2] H. C. Robbins Landon: Haydn. Chronicle and Works. Vol. 1, Thames and Hudson 1980, p. 563.

[3] The wind parts, at this point, have piano in square brackets in the Robbins Landon edition, but it would surely be an excess of musicological purity to take the lack of a marking by Haydn to mean that they, and they alone, should continue forte.

[4] Ibid.

[5] Robbins Landon (ibid) states firmly that Haydn, after the piano beginning of the recapitulation, reserves “the forte for a dramatic fanfare for the four horns, striding up the open harmonics of their instruments”. However, in Robbins Landon’s edition, the forte marking for the horns is in square brackets, indicating that it is editorial. Only the in the violin parts is the forte without square brackets, and therefore Haydn’s own. All logic would seem to side with Robbins Landon’s assumption that the forte automatically applies to all instruments, but pure musicology might leave space to disagree. Leinsdorf has the horns continue piano, while Märzendorfer has them play the passage in crescendo, beginning piano.

[6] https://archives.bso.org/

[7] I have not found a review of the March performance. This review, in The Day of 11 April 1967, is of a repeat in the Palmer Auditorium.

[8] Chicago Tribune 7 April 1967. The BSO were guests in Symphony Hall, Chicago while the CSO was on tour.

[9] The cello soloist is not named in the programme booklet, but the Chicago notice (above) names Eskin, who was principal cellist of the orchestra from 1964 to 2016 and so presumably played at all the performances.