Unmistakably, Mozart takes singing as his starting-point, and from this issues the uninterrupted melodiousness which shimmers through his compositions like the lovely forms of a woman through the folds of a thin dress.
Let this be the first warning to the Mozart performer: piano playing, be it ever so faultless, must not be considered sufficient. Mozart's piano works should be for the player a receptacle full of latent musical possibilities which often go far beyond the purely pianistic. It is not the limitations of Mozart's pianoforte (which I refuse to accept) that point the way, but rather Mozart's dynamism, colourfulness and expressiveness in operatic singing, in the orchestra, in ensembles of all kinds. For example, the first movement of Mozart's Sonata in A minor K 310 is to me a piece for symphony orchestra; the second movement resembles a vocal scene with a dramatic middle seciton, and the finale could be transcribed into a wind divertimento with no trouble at all.
In Mozart's piano concertos, the sound of the piano is set off more sharply against that of the orchestra. Here the human voice and the orchestra solo instrument will be the main setters of standards for the pianist. From the Mozart singer he will learn not only to sing but also to 'speak' clearly and with meaning, to characterize, to act and react; from the string player to think in terms of up-bow and down-bow; and from the flautist or oboist to shape fast passages in a variety of articulations, instead of delivering them up to an automatic non-legato or, worse still, to an undeviating legato such as the old complete edition prescribed time and again without a shred of authenticity.
A singing line and sensuous beauty, important as they may be in Mozart, are not, however, the sole sources of bliss. To tie Mozart to a few traits is to diminish him. That great composers have manifold things to say and can use contradictions to their advantage should be evident in performances of his music. There has been altogether too much readiness to reduce Mozart to Schumann's 'floating Greek gracefulness' or Wagner's 'genius of light and love'. Finding a balance between freshness and urbanity ('He did not remain simple and did not grow over-refined,' said Busoni), force and transparency, unaffectedness and irony, aloofness and intimacy, between freedom and set patterns, passion and grace, abandonment and style -- among the labours of the Mozart player, this is only rewarded by a stroke of good luck.
What is it that marks out Mozart's music? An attempt to draw a dividing-line between Haydn and Mozart could perhaps help to answer the question. Mozart sometimes comes astonishingly close to Haydn, and Haydn to Mozart, and they shared their musical accomplishments in brotherly fashion; but they were fundamentally different in nature. I see in Haydn and Mozart the antithesis between instrumental and vocal, motif and melody, C. P. E. and J. C. Bach, adagio and andante, caesuras (amusing and startling) and connections (seamless), daring and balance, the surprise of the unexpected and the surprise of the expected. From tranquillity, Haydn plunges deep into agitation, while Mozart does the reverse, aiming at tranquillity from nervousness.
Mozart's nervous energy -- his fingers were constantly drumming on the nearest chair-back -- can be recognized in the fidgety or spirited agitation of many final movements, as one heard them in performances by Edwin Fischer, Bruno Walter or Artur Schnabel. When Busoni denies Mozart any nervousness, I have to disagree. Like melodiousness shimmering through the folds of a dress, 'chaos' now and then, even in Mozart, can be 'shimmering through the veil of order' (Novalis).
The perfection of that order, the security of Mozart's sense of form, is, as Busoni puts it, 'almost inhuman'. Let us therefore never lose sight of the humanity of this music, even when it gives itself an official and general air. The unimpeachability of his form is always balanced by the palpability of his sound, the miracle of his sound mixtures, the resoluteness of his energy, the living spirit, the heartbeat, the unsentimental warmth of his feeling.
Between Haydn the explorer and adventurer, and Schubert the sleepwalker, I see both Mozart and Beethoven as architects. But how differently they built! From the beginning of a piece, Beethoven places stone upon stone, constructing and justifying his edifice as it were in accordance with the laws of statics. Mozart, on the other hand, prefers to join together the most wonderful melodic ideas as prefabricated components; observe how in the first movement of K. 271 he varies the succession of his building-blocks, to the extent of shaking them up as though in a kaleidoscope. Whereas Beethoven draws one element from another, in what might be called a procedural manner, Mozart arranges one element after another as though it could not be otherwise.
Mozart, more than most other composers, expresses himself differently in minor and in major keys. That he could also compose in a procedural manner is demonstrated by his two concertos in minor keys, K. 466 and 491, which so greatly impressed Beethoven. Original cadenzas for these two works unfortunately do not exist. Neither the dynamic spaciousness of the D minor concerto nor the contrapuntal density of the C minor concerto is compatible with the usual type of improvisational cadenza in Mozart's concertos in major keys. Rather more conceivable are cadenzas in the manner of Bach's Fifth Brandenburg Concerto which carry on the intensity of the movement, transporting it in a broad arc to the next entrance of the orchestra.
Mozart is made neither of porcelain, nor of marble, nor of sugar. The cute Mozart, the perfumed Mozart, the permanently ecstatic Mozart, the 'touch-me-not' Mozart, the sentimentally bloated Mozart must all be avoided. There should be some slight doubt, too, about a Mozart who is incessantly 'poetic'. 'Poetic' players may find themselves sitting in a hothouse in which no fresh air can enter; you want to come and open the windows. Let poetry be the spice, not the main course. It is significant that there are only 'poets of the keyboard'; a relatively prosaic instrument needs to be transformed, bewitched. Violinists, conductors, even lieder singers -- so usage would suggest -- seem to survive without 'poetry'.
One look at the solo parts of Mozart's piano concertos should be enough to show the Mozart player that his warrant leaves that of a museum curator far behind. Mozart's notation is not complete. Not only do the solo parts lack dynamic marking almost entirely; the very notes to be played -- at any rate in the later works that were not made ready for the engraver -- require piecing out at times: by filling (when Mozart's manuscript is limited to sketchy indications); by variants (when relatively simple themes return several times without Mozart varying them himself); by embellishments (when the player is entrusted with a melodic outline to decorate); by re-entry fermatas (which are on the dominant and must be connected to the subsequent tonic); and by cadenzas (which lead from the six-four chord in quasi-improvisational fashion to the concluding tutti).
Luckily, there are a good number of Mozart's own variants, embellishments, re-entries and cadenzas, and they give the player a clear idea of his freedom of movement. In re-entries and cadenzas the main key is never deviated from; in embellishments and variants the prevailing character is never disturbed. Mozart's variants sometimes show a subtle economy which, I assume, was not in keeping with contemporary convention. (In his C minor Concerto K. 491 the extremely delicate shifts of harmony, part-writing and rhythm at the returns of the initial theme should be savoured without further additions.) The view that empty spots must stay empty because the performer cannot possibly claim to possess Mozart's genius has been overcome today; it was an attitude produced by misguided reverence, which did not expect or trust the player to have the necessary empathy with Mozart's style. The case of the Rondo in A major K. 386 is instructive; thanks to the recent discovery of the last pages of Mozart's manuscript, we now realize that the final twenty-eight bars of the Rondo, as we used to know it, are not by Mozart but by Cipriani Potter, which no one would otherwise have noticed.
It is precisely in those passages where Mozart's text is sketchy that the player must know exactly what Mozart wrote and how he wrote it, but not put his faith in editors. Anyone who takes on Mozart's piano concertos will have to devote some time to studying the sources. A particular case in point is the so-called 'Coronation' Concerto K. 537. Most of the left hand is not worked out at all. In the middle movement, which is plagued by a complete lack of emotional contrast, the same four-bar phrase appears no fewer than ten times in virtually identical guise. Here the richest ornamentation will be needed if the effect is not to resemble the pallid charm of certain Raphael Madonnas, which the nineteenth century adored, just as it did this movement, unembellished. It is not at all easy to understand why a version of this lovely work fabricated after Mozart's death is still generally played today, as though nothing about it could stand to be improved.
Additions to Mozart's text are in some instances obviously required, in others at least possible. An appendix to the Bärenreiter complete edition prints a lavishly embellished version of the F sharp minor Adagio from the Concerto in A major K. 488; it is probably the work of a pupil, and apparently was part of Mozart's musical estate. What is elaborated in this manuscript is in no way satisfactory, but it does provide a clue that embellishment is permitted. As to how one is to go about it, Mozart's own models, and no others, are the ones to be guided by. The embellishments by Hummel or Philipp Karl Hoffmann do not even try to follow Mozart's example; they are foreign to his style and frequently overcrowded with notes to such a degree that, to get all of them in, the relatively flowing tempi or Mozart's middle movements must be pulled back to largo. The additions by Hummel and Hoffmann do make us aware that the 'gusto' of performance style could change quite quickly and drastically; this should give pause to those who try to get at Mozart by concentrating too single-mindedly on baroque practice.
The player's delight at filling in the white spots on Mozart's musical map in such a way that even the educated listener does not prick up his ears must stay within bounds. The player must not be seduced into overdoing it or into living too much for the moment. When improvising embellishments becomes a parlour game gleefully played to flummox the orchestra, when the player sets out in every performance to prove to himself and all present that he is indeed spontaneous he is in danger of losing control over quality. I think he will be more deserving if he makes a rigorous selection from a supply of versions he has improvised at home, rather than risking everything on the platform by trying to play Mozart as though he were Mozart.
One of the additions that is possible but rarely necessary, since in most cases it merely doubles the orchestra, is continuo playing. Once I relished accompanying the bass line of the orchestra, but today I usually limit myself to taking a hand occasionally in energetic passages and to giving almost imperceptible harmonic support to some piano cantilenas. At a time when there were neither conductors nor full scores, the basso continuo, apart from giving the soloist his harmonic bearings, served mainly to co-ordinate the players' rhythm. Nowadays one can reasonably expect the soloist to be familiar with the score (lately even lieder singers are expected to have taken a glance at the piano accompaniment); and naturally we expect the conductor to keep the orchestra together. Basso continuo playing therefore seems to have a point only in special cases, such as when the four Mozart chamber concertos (K. 413-415 and 449) are performed without winds. But the difference between solo and tutti must not be lost.
Even a composer like Mozart could make a mistake. Artur Schnabel's precept that the performer must accept the whims of great composers though he may be quite unable to fathom them must not be allowed to go so far that errors remain unrectified. Schnabel himself provided some examples of reverential blindness, as when, for example, in the middle movement of the Concerto in C minor K. 491, he played a bar, with wind accompaniment, precisely as Mozart inadvertently let it stand. Here, as in one bar of the finale of K. 503, Mozart apparently wrote the piano part first and then, when writing in the orchestral parts, changed his mind about the harmony. In doing so he forgot to adjust the piano part to the new harmonic situation. The result is cacophony and a divergence in the leading of the bass line that is unthinkable in Mozart. If the player, in rare instances, puts Mozart's text right, it does not mean that he presumes himself to be equal, or indeed superior, to Mozart.
With the alla breve of the middle movement of K. 491, Mozart seems to set us a riddle, but for once without giving us 'the solution with the riddle' (to quote another of Busoni's Mozart aphorisms). Paul and Eva Badura-Skoda have gone to some lengths to explain why Mozart must have made a mistake with this marking. In its note values, the movement is twice as slow as the alla breve movements in Concertos K. 466, 537 and 595. As confirmed by the textbooks of the period, and by Beethoven's metronome figures, the alla breve marking stands not only for counting half-bars but also for considerable increase in tempo. Yet there are exceptions, as Erich Leinsdorf has been kind enough to point out to me, and the second movement of K. 491 is one. Leinsdorf mentions, among others, some examples from The Magic Flute (Overture: Adagio; No. 8: Larghetto; Act II: March of the Priests; No. 18: Chorus of the Priests; No. 21: Andante) where the alla breve 'should be translated to a contemporary conductor meaning: in four, my boy, not in eight'. But there is also the Aria with gamba 'Es ist vollbracht' ('It is finished') from Bach's St John Passion where Bach indicated, above the 3/4 of the middle section, the words alla breve, suggesting the 'next faster unit': in three, not in six. The old complete edition, which altered several of Mozart's tempo markings arbitrarily, transformed the alla breve in the first movement of the Concerto in F major K. 459 into 4/4 time, thereby doing precisely what this piece cannot tolerate: it is meant to move along not alla marcia, as we are constantly told in commentaries and hear in performances, but dancingly and in whole bars.
Mozart was not a flower child. His rhythm is neither weak nor vague. Even the tiniest, softest tone has backbone. Mozart may dream now and then, but his rhythm stays awake. Let the tempo modifications in Mozart be signs of a rhythmic strength that counterbalances emotional strength; above all in variation movements, it will surely be permissible to graduate the tempo at times, to set off the variations from one another. Mozart may lament -- and that lamentation can reach a pitch of solitary grief -- but he does not moan and groan. Two-note patterns should be 'sighed' only when the music really demands sighing. Not only singers should be aware of the difference between a suspension, which has a purely musical role, and an appoggiatura, whose role is emotional and declamatory, stressing the pathos of two-syllable words.
Is Mozart's music simple? For his contemporaries it was frequently too complicated. The idea of simplicity has become downright embarrassing in this century. There is a 'kitsch' of plainness, especially noticeable in the literary glorification of the 'simple life' and in a longing for the 'popular vein'. What was all right for the Romantics is thought to be reasonable enough for their descendants. Simplicity in playing Mozart must not mean subjecting diversity to a levelling process or running away from problems. Simplicity is welcome as long as the point is to avoid superfluity. But to 'concentrate only on what counts' in Mozart is questionable. Everything in his music counts, if we leave out a few weaker works or movements, of which there are some even among Mozart's piano concertos, for example the early pieces preceding that wonder of the world, the 'Jeunehomme' Concerto K. 271.
The identity of Mlle Jeunehomme seems to remain just as mysterious as the sudden supreme mastery that unfolds in the work composed for her. Here it is revealed for the first time that Mozart is both 'as young as a youth' and 'as wise as an old man' (Busoni). And from this point on, the Mozart player must shoulder a burden of perfection that goes beyond his powers.