Why do some people seem to have a problem with Mozart’s Piano Concerto in D, K537? Since the Second World War the reputation of Mozart’s penultimate piano concerto seems to have nosedived while that of its sister works has held firm. True, it may not boast the Sturm und Drang drama of the D minor or C minor Concertos (K466 and 491), the piercing elegance of the Elvira Madigan (K467), the playfulness of the A major (K488), the grandeur of the E flat (K482) or C major (K503), or what some hear as the valedictory qualities of the last concerto (K595). Nevertheless, K537 remained popular throughout the 19th century and into the 20th; Friedrich Blume wrote in 1935, in the preface to the Eulenburg edition, that it was ‘the best known and most frequently played’ of Mozart’s concertos; it was the second to be recorded, after Dohnányi’s G major (K453) in 1928, and was second in popularity on disc during the 78 era only to the D minor.

Things changed suddenly after the war. In 1948 Cuthbert Girdlestone decried its ‘impertinent and irrelevant virtuosity’, pronounced the slow movement ‘neither great nor deep’ and found the finale ‘even slighter than the first movement’, while Arthur Hutchings deemed it sad proof that ‘Mozart’s last years were not in tune with the concerto form’. Quoting the melody in bars 44-48 of the Larghetto, he wailed that ‘one can only regret that Mozart stooped so low’. (He didn’t express the same reservations over bars 40-51 of K466’s Romance or bars 49-58 of K595’s Larghetto, which after all perform an almost identical harmonic function with similar gestures.) Subsequently the Coronation’s esteem among pianists has suffered, and it appears on recordings with considerably less frequency than the surrounding concertos, K503 and K595.

There is a range of possible reasons for the work’s comparative disregard, bound up with its genesis and early performance history. Mozart entered it in the Verzeichnüss aller meiner Werke – the catalogue he kept of his works from 1784 onwards – on 24 February 1788, although paper studies have shown that it may have been begun as early as the beginning of 1787. That year the composer was kept busy with preparations for Don Giovanni, premiered in Prague in October, and with the fallout from the death of his father in May. Mozart presumably returned to the concerto fragment in advance of a proposed Akademie in Vienna in 1788 but there is no firm evidence of any such concert; he gave the first known performance in Dresden on 14 April 1789 before the Elector of Saxony, Friedrich August III, and his wife Amalie. The work owes its common nickname, Coronation, to a performance on 15 October the following year at the celebrations in Frankfurt connected to the crowning of the new Holy Roman Emperor: when the concerto was published in 1794 by JA André, the title-page bore the inscription ‘Ce Concerto a été executé par l’Auteur à Francfort sur le Mein à l’occasion du Couronnement de l’Empereur Leopold II’.

The autograph reveals that however protracted the concerto’s period of composition, it was assembled hurriedly, in bursts of activity. Mozart appears to have embarked on a work for piano and strings alone before adding woodwind parts that nevertheless do not boast the prominence they are accorded in his other Viennese piano concertos. The addition of trumpets and timpani, too, appears to have been an afterthought: these instruments are entered in the score in their correct position only from bar 173 of the finale, before Mozart went back and inserted their parts on the top and bottom staves of the manuscript up to that point. Moreover, his description of the work in the Verzeichnüss indicates that the wind, brass and timpani were ad libitum – that they could be omitted in performance.

Further evidence of haste is found in the piano part, which is far from complete in the manuscript. The right hand is entered throughout but the left hand is absent for much of the outer movements and the whole of the Larghetto. Mozart, of course, knew how the missing parts would go in performance but never got round to writing them down. André either composed or commissioned a completion of the piano part, which duly appeared in his 1794 edition and became the bedrock of K537’s performing tradition, although these insertions tend to be efficient rather than imaginative – largely offering little more than chordal or Alberti-style accompaniment – and an increasing number of pianists have felt no compunction in providing their own solutions where they feel they can improve upon André’s. There are plenty of points where the given accompaniment seems no more than workaday, or where a sparse left hand leaves a melody naked and unsupported. Some harmonies would benefit from discreet filling out, and there’s even a bar towards the end of the Larghetto (bar 105) where the top line simply drops out – as if Mozart was distracted. While most pianists freely ornament the cantilena of the central movement, it’s surprising how many offer no fig leaf for this solecism.

Much of this may have led to the view that K537 is a poor relation to the other late concertos. Nevertheless, there are signs of a backlash to this assumption. Simon Keefe, in the The Cambridge Mozart Encyclopedia (CUP: 2006), describes K503 and 537 as ‘fundamentally misunderstood works’ that ‘bear witness … to an active reappraisal of [the piano concerto’s] stylistic features’. It was a favourite of Friedrich Gulda, Alicia de Larrocha and Maria João Pires, while pianists of the current generation such as Francesco Piemontesi have seized upon the work for recording projects ahead of its grander siblings. Mitsuko Uchida credits Colin Davis with deepening her appreciation of the work, telling Classic FM’s website: ‘He thinks it’s an underrated masterpiece. Increasingly I think he’s right.’ (Not that it was among the works chosen when she revisited a selection of the concertos with the Cleveland Orchestra.) Jean-Efflam Bavouzet, too, is open about his change of opinion as a result of prolonged acquaintance with the concerto. It is, after all, a work by Mozart from the mid- and late 1780s – a period during which he was at the height of his powers and barely wrote a bad note.

The earliest recordings

Magda Tagliaferro was the first to take the work into the studio, in 1930. Her sensitive pianism is supported by Reynaldo Hahn’s rather rough Pasdeloup Orchestra, presenting a slightly abridged version (with cuts in the opening tutti and the finale). Robert Casadesus followed only a year later, although the performance style and recording quality could be from a different world. Walther Straram’s orchestra betray a lack of rehearsal as the work proceeds; episodes are raced through with little delineation of character and the Larghetto is filleted down to little over two minutes! Wanda Landowska, in a recording made to present to King George VI in 1937, shapes and shades the fast movements well, although the Larghetto wilts under a slow tempo and indulgent rubato. Gina Bachauer is an undemonstrative but thoughtful soloist, a little careful in the finale, where the New London Orchestra’s accompaniment becomes somewhat ordinary.

The best of the 78s, though, comes from Wilhelm Backhaus in a wartime recording made to mark the 150th anniversary of Mozart’s death. Uniquely for the period, Backhaus, with full-blooded piano tone and more pronounced use of the pedal, augments André’s contributions to the piano part, much to the music’s advantage. His Larghetto is on the slow side but beautifully sustained and ornamented, and he improvises an ambitious Eingang in the finale, playing around with the rising scales from the introduction to the finale of Beethoven’s First Symphony.

Into the LP age

Leave aside a sluggish 1952 taping on which an unsympathetic Carl Seemann plays alongside Fritz Lehmann’s uningratiating Berlin Philharmonic in favour of Friedrich Gulda’s 1955 recording. The combination of Gulda’s wide-ranging imagination and the Mozartian experience of Anthony Collins is a potent one, let down only by a rather veiled piano tone (in the Carlton Rooms at Maida Vale). Gulda is all improvisatory freedom and restores a touch of devilry to the finale that he would recreate three decades later within the far grander orchestral conception of Nikolaus Harnoncourt with the Concertgebouw Orchestra, along with a mesmerising, vocally phrased Larghetto. Then again, with the Munich Philharmonic in 1986 he is truly among friends and they play the hell out of the work in a reading far removed from notions of genteel historic style – brash piano sound, moments of unsteadiness, odd missed notes and all. You may find it overdone or even vulgar but it grips from first to last.

Casadesus provided a full text for the central movement of his 1962 stereo remake, equalled in elegance a year earlier by the slightly more recessed piano sound (at Abbey Road) of Hans Richter-Haaser, even if István Kertész’s Philharmonia are no match for the sophisticated refinement of George Szell’s Columbia SO for the Frenchman. Géza Anda’s 1965 recording was damned with faint praise in a contemporary review for not attempting ‘more than fluency and prettiness’, although his piano is caught beautifully by DG and the Salzburg Mozarteum orchestra are stylish and responsive under the pianist’s direction. On the other hand, the Vienna Festival Orchestra under Stephen Simon are brasher for Lili Kraus the same year – she reportedly beseeched them continually to play ever quieter – and her particular hold-backs and desynchronisations lead to the solo part becoming momentarily unmoored from the orchestra in places. Her Larghetto is refreshingly unsentimental but her finale unsmiling in comparison to Anda’s swifter, playful levity.

Kraus’s Mozart concertos did not receive an official UK release until 2017, and Clifford Curzon’s 1967 Coronation with Kertész’s beefy LSO was similarly disinterred only in 2002. A couple of years earlier BBC Legends issued a Curzon performance from a BBC Prom on 14 August 1974 that boasts all the personality, atmosphere and inner illumination that seem missing from the studio version. The conductor of the concert, perhaps a little incongruously, was Pierre Boulez, who sandwiched the Mozart between more characteristic repertoire: Berio’s Allelujah II and Bartók’s Concerto for Orchestra. Given the infamous Albert Hall acoustic and a fair few finger-fluffs, it’s no library choice but perhaps explains the reason for withholding the comparatively buttoned-up studio recording. And in his exquisitely ornamented Larghetto, Curzon beguiles the bronchial Proms audience briefly into awed silence.

Ingrid Haebler is another repeat visitor to K537. Her mid-1950s recording with the Vienna Pro Musica Orchestra and Heinrich Hollreiser was superseded and outclassed by her 1961 reading with Colin Davis and the LSO, which is itself less readily accessible now than her LSO remake with Witold Rowicki. Her fastidious and cultivated pianism nevertheless operates within a frustratingly circumscribed emotional and dynamic range; Rowicki has finer engineering but lacks Davis’s deep affinity with Mozart.

Seventies and Eighties cycles

The period from the late 1960s saw the initiation or continuation of a number of concerto cycles (or near cycles) by high-profile pianists – most of them approaching the Coronation later rather than earlier in their endeavours. Daniel Barenboim demonstrates aristocratic pianism and a natural instinct to integrate with his orchestral colleagues. His interpretation doesn’t alter much between his 1973 recording as part of the cycle he made directing the English Chamber Orchestra from the keyboard and his 1989 remake with the Berlin Philharmonic; the latter orchestra is rather more finely disciplined, although Barenboim’s dexterity isn’t quite so infallible second time around. The Larghetto is the emotional core of both performances, magnificently sustained and discreetly ornamented.

Conversely, Maria João Pires’s 1974 Lisbon recording is simply outclassed by her return to the work in concert with Claudio Abbado’s Vienna Philharmonic in 1990. The Austrian players could give their Gulbenkian counterparts a lesson or two in refinement, and the Portuguese piano is a rather clunky, harshly recorded instrument. But Pires is clearly inspired by the live situation at the Salzburg Festival, and turns in an imaginatively decorated Larghetto and an engagingly genial finale imbued with the spirit of the dance.

Vladimir Ashkenazy offers a finely turned but not particularly penetrating 1983 performance, directing an efficient rather than ardent Philharmonia from the keyboard. Just a few years earlier, with the same orchestra, Tamás Vásáry made a rare foray into 18th-century repertoire and proved himself a touch more responsive, inviting a more classy accompaniment and offering a tastefully embellished Larghetto.

The cycles by Alfred Brendel and Murray Perahia emerged almost simultaneously and represent strikingly different approaches to Mozart by two of the leading pianists of the time: the Dionysian Brendel and the Apollonian Perahia. Where Brendel sees the score as a provocation, Perahia presents unruffled, classically poised performances of the concertos. Both cycles were welcomed as they gradually appeared, while critics were quick to contrast Brendel’s range of touches and dynamic shades with the grace and intuitive beauty of Perahia’s performances. Their Coronations were both recorded in 1983 and typify the styles of the two musicians. Both amply augment Mozart’s bare piano parts, most notably in the Larghetto, Perahia here providing a songful cantabile but Brendel creating something almost operatic from the innocent melody. And while both finales are joyously effervescent, Brendel’s molten surface occasionally runs away with itself in the passagework. Despite (or perhaps because of) their stark differences, the two collected cycles set standards for Mozart performance in the early CD era: Mozart contains multitudes.

The cycle by Mitsuko Uchida with the English Chamber Orchestra under Jeffrey Tate also became a benchmark but has tended to divide opinion still further. There is no denying Uchida’s deep and genuine understanding of Mozartian style, nor her exquisitely weighted fingerwork and the balance between piano and orchestra that she and Tate worked so hard to perfect. Against this, though, there is a tendency to draw back from moments of drama, bypassing the turbulence that so often enriches the late concertos, and to indulge in a super-delicate pianissimo that draws the attention, to be sure, but imposes a fragility upon music that is in reality more robust, however elegant and graceful it appears. This somewhat neuters the Larghetto, which is already among the slower of those considered here, while the dynamic range of the finale hardly ever peeps over mezzo-forte and comes across as fatally precious and mannered. ‘There seems an excess of attention to the less important things, in … slow movements particularly, which are flawed by an overlay of artifice – well-meaning and charmingly applied but tending to turn every statement into something exquisitely fragile,’ wrote Stephen Plaistow in his November 1988 Gramophone review of the original release of Uchida’s K537 and 595. ‘Nothing is straightforward because nothing is left untouched.’

Alicia de Larrocha recorded a sequence of 10 predominantly late piano concertos with the English Chamber Orchestra under Colin Davis between 1991 and 1993 but a superior 1985 version of the Coronation with Georg Solti’s Chamber Orchestra of Europe was prevented from release at the time by concerns expressed by both pianist and conductor over the boomy acoustic of Henry Wood Hall. Its remastering and release 25 years later reveals Larrocha’s beautiful touch in this music, her pianism more assured than in the remake, embedded more naturally in its dialogue with Solti’s superior accompaniment.

Jeno˝ Jandó rather unfairly gained a reputation as something of a workhorse for Naxos during its early years, ploughing through an astonishing range of challenging repertoire at an almost industrial rate. His Coronation, part of a concerto cycle made to coincide with the Mozart bicentenary commemorations of 1991, nevertheless demonstrates his natural affinity with 18th-century style and his alertness to the dramatic contours of the work, enhanced by the presence of a chamber orchestra, Concentus Hungaricus, that can confidently hold its own against starrier Western bands.

András Schiff belatedly topped out his concerto cycle with the Coronation in 1993. He is one of the first pianists of the modern era to take a dissenting view on the standard realisation of the missing left-hand parts, providing his own solutions in places where André’s seem simplistic or harmonically incomplete. In the slow movement, for example, he eschews the left hand’s echo of the right in bar 2, notwithstanding the heavy hint from the woodwind a few bars later that it is correct. Schiff deploys a full range of dynamics, the slightly rounded attack of his Bösendorfer suiting his essentially lyrical conception of the work. Affected pianissimos in the finale, though, feel perverse in a performance that otherwise bubbles and froths but never actually sounds much fun.

Period-instrument performances

As early as 1970, Jörg Demus recorded a handful of concertos with Collegium Aureum. The small, silvery tone of his fortepiano (a Viennese instrument from c1790 by Schantz) comes as a surprise after the Coronation’s robust orchestral introduction but the instrument is never quite submerged by the band, possibly thanks more to canny microphone placement than instrumental balance. The Larghetto operates at two markedly different tempos but is expressed simply and largely without ornamentation, save for use of the damper. Some deliciously unruly tuning in the woodwind and brass is a symptom of audible fatigue from the band in the closing minutes but this recording serves as a stimulating reminder of how ‘authenticity’ was done more than half a century ago.

Four period-instrument recordings made since the advent of CD use replicas of keyboards by Mozart’s preferred maker, Anton Walter – with the exception of Robert Levin, who made the first concerto recording on Mozart’s own Walter from c1782. Nevertheless, the first of this group of recordings was made in 1986 by Malcolm Bilson with John Eliot Gardiner’s English Baroque Soloists, displaying higher playing standards than could be approached by Collegium Aureum two and a half decades earlier. The blend between piano and orchestra, too, is more natural, although the dynamic range is somewhat narrow.

Jos van Immerseel five years later also operates within a constrained dynamic spectrum in a performance with Anima Eterna (making their recording debut) that feels rather cosmetic, with a repertoire of ornamentations and decorations that doesn’t seem entirely natural or stylistic. The lighter-hearted finale is the most successful movement, although the placing in the sound picture of the piano among the orchestra leads to its filigree often being swamped.

Robert Levin reached the Coronation in 1997, six years into his cycle with the Academy of Ancient Music and Christopher Hogwood but only a year before Decca prematurely pulled the plug on the project. Levin not only deploys a greater variety of dynamic and rhythmic inflection than Bilson but also joins in during tuttis like so few of his predecessors and, most importantly, provides his own realisation of much of the missing left-hand part from as early as his first solo entry: passages such as the sparely harmonised chromatic figures in bars 133-31 and 134-35 and their returns, for example, are filled in more imaginatively. The Larghetto is perhaps a little over-decorated but Levin’s lifetime immersion in Mozart’s music ensures that he never transcends the bounds of 18th-century style. By contrast, Ronald Brautigam’s reading appears to skate across the surface of the music, his first two movements the swiftest among the recordings considered here, the Larghetto coming over as especially facile.

Recordings since 2000

Matthias Kirschnereit and Vassily Primakov both embarked on concerto series that show them to be perceptive and stylistically aware Mozartians, although in performances that are not at the level of the finest. Kirschnereit, whose Bamberg cycle was subsequently made available in a 10-disc box-set, sounds slightly impatient in the Larghetto and rather unrelenting in the finale’s passagework, with imperfect balance between piano and orchestra. Primakov (nine concertos recorded in Odense, 2008-11) is spotlit in relation to the orchestra, favouring brisk proficiency and responsive shaping. His decorations in the Larghetto include playing the final iteration of the opening melody an octave higher than written, while his finale fizzes away at the expense of absolute security and characterisation.

The two most recent recordings in this selection bear a number of similarities: both Francesco Piemontesi and Jean-Efflam Bavouzet work with chamber orchestras – modern instruments with period manners – and share approaches not only to tempo (with swift slow movements) but also to reimagining the inauthentic left-hand line. Piemontesi comes up with the most persuasive solution yet for the filling-in of bars 130 passim and alters the harmony from bar 31 in the Larghetto. Bavouzet’s interventions, too, enrich harmonies from bar 96 in the opening movement and bar 82 of the Larghetto. The vibrato-light strings of the Scottish Chamber Orchestra and Manchester Camerata may not be to all tastes but both are under the direction of musicians – Andrew Manze and Gábor Takács-Nagy – who respectively bring period awareness and chamber attentiveness to their accompaniments.

Cadenzas and a final flourish

Unsurprisingly, given the incompleteness of the manuscript of K537, it comes down to us without an authentic cadenza. Matters were complicated by the publication by Breitkopf of an edition that mistakenly assigned an authentic cadenza for the earlier D major Concerto, K451, to the Coronation: it fits rather well, in fact, and was the choice of Casadesus, Bachauer, Richter-Haaser, Pires (in her first recording) and Larrocha. Most other pianists concoct their own cadenza: Landowska’s presses into service themes from The Marriage of Figaro and was later adopted by Curzon, Barenboim and Primakov. The current favourite, though, appears to be one by Paul Badura-Skoda, to which Pires switched for her second recording and which was selected by Piemontesi and Bavouzet.

Christian Zacharias, like most of the other pianists in this selection, composed his own cadenza for his two recordings. Both are highly personal readings, and in the earlier of the two he even contrives additional woodwind parts for the Larghetto. The cadenza comes as the biggest surprise in both recordings: part of it is played not on the piano but on what he described as ‘a Japanese synthesiser toy’, giving a music-box effect. An individual touch, to be sure, and one that caught the ear of Francesco Piemontesi, who borrowed Zacharias’s cadenza for his 2015 BBC Proms performance with the Aurora Orchestra, conductor Nicholas Collon tinkling away on a celesta.

Piemontesi is one of a new generation of pianists espousing a more positive opinion of K537 than has prevailed over recent decades, along with an intelligent, imaginative and innovative approach to the partial piano part. It’s surprising, perhaps, that so few pianists take such a reconstructive approach, or that no musicologist has yet come up with a more persuasive realisation – although the American pianist-composer Timo Andres offers a startling ‘recompositon’ (Nonesuch), filling in the gaps in his own postmodern manner.

As for a preference among these 40 recordings, it is the ones that take a more proactive approach to the lacunae of Mozart’s manuscript that tend to appeal most. ‘How difficult it is to add anything at all to Mozart’s text,’ wrote Stephen Plaistow. ‘Embellishments in the concertos, however necessary, have a way of never sounding exactly what Mozart himself would have done in the circumstances.’ Brendel perhaps comes closest; it must remain a matter of personal taste whether Levin, Piemontesi and Bavouzet go too far.

Maria João Pires is more restrained in her embellishment of the Larghetto’s cantilena, and offers still subtler tweaks to the traditional score in the outer movements. The presence of an audience and the retention of applause on the recording may be a deal-breaker for some but it is Pires – with her kaleidoscopic range of touch, her inability to leave a string of semiquavers uninflected and her sense of when to leave well alone – whose recording lingers most firmly in the memory. ‘Exploring with sensitivity the considerable areas of sadness and loneliness … is to dispel the myth of the “heartlessness” and “coldness” of the much undervalued Coronation Concerto,’ wrote Deryck Cooke in 1965. In her vivid response to the Allegro’s withdrawal into contrapuntal contemplation or the finale’s repeated retreat to minor tonalities, Pires remains the ideal advocate for this under-appreciated but endlessly fascinating concerto. IP

With thanks to Mike Spring for his assistance with details of the earliest recordings

This feature originally appeared in the SPRING 2025 issue of International Piano