I must begin with a qualification: this first recording of Beethoven's piano works, which I made for Vox-Turnabout between 1958 and 1964, is not entirely complete. There seemed to me little virtue in rescuing from oblivion works that are totally devoid of any touch of Beethoven's mastery and originality. It was without regret, therefore, that I omitted pieces like the deplorable Haibel Variations, which could have been written by any of Beethoven's contemporaries, as well as certain student exercises, Albumblätter, studies, sketches and curiosities, most of which were never intended for publication -- pieces, that is, which are merely of interest to the historian. These include the total output of the Bonn period (among which are the Variations on a March by Dressler by the twelve-year-old Beethoven and the two preludes through all the major keys, curiously published later on as Op. 39), the Easy Sonata in C major WoO 51, the Variations on the 'Menuet à la Vigano' by Haibel which I have already mentioned, the pieces WoO 52, 53, 55 (the Prelude in the style of Bach), 56, 61 and 61a, as well as the little dance movements WoO 81-86, of which I retained only the Six Écossaises, WoO 83, although in all likelihood these are transcriptions of an orchestral score, and the single extant copy, passed down by Nottebohm, may well be dubious in some of its detail. It is not for nothing that virtuosi have been stimulated again and again to make arrangements of these spirited pieces.

If I mention the fact that I concluded the series at the age of thirty-four, this is not to plead for mitigation, but to acquaint the reader with a circumstance that may explain certain features of these interpretations. Nothing was further from my mind than to suppose that I could present in my recordings anything like a definitive solution of the Beethoven problem. Nor was it my intention to supply the musical illustrations to any fashionable theory of Beethoven interpretation. I just plunged into an adventure, the consequences of which I could no more foresee than could the record company that had put its trust in me.

My work on the Beethoven series took five and a half years. One of the crosses the artist has to bear is that the date of a recording is so rarely indicated on the record sleeve. He is all too easily blamed or, almost worse, praised for interpretations that have lost some of their validity, at least as far as he himself is concerned. People expect an artist to develop, and yet they are only too ready to impale him, like an insect, on one of his renderings. The artist should have the right to identify his work with a certain phase of his development. It is only the continuous renewal of his vision -- either in the form of evolution or of rediscovery -- that can keep his music-making young.

The recordings of Beethoven's variation works, with the exception of the Diabelli Variations, were made in three stages between December 1958 and July 1960. There followed, at the turn of 1960/61, the last five sonatas, together with the Fantasy, Op. 77. In March 1962 I played the Sonatas Op. 31 Nos.1 and 2, Op. 57 and Op. 90; in June and July of that year all the remaining sonatas between Op. 22 and Op. 81a. The early sonatas from op. 2 to Op. 14 were recorded in December 1962 and January 1963. (By coincidence, I concluded my work on the thirty-two sonatas on my thirty-second birthday.) Finally, in July 1964, I played the miscellaneous pieces and the greatest of all piano works: the Diabelli Variations.

I recall a cold winter morning in a rather dilapidated Baroque mansion in Vienna; the logs in the fire-place of the hall where we recorded crackled so loudly that we had to throw them out of the window into the snow. Several changes in recording technique, and in the room and instrument, proved unavoidable. In the event, there were five groups of recordings: 1) the variation works, 2) the late sonatas, 3) the middle-period sonatas from Op. 22 on, 4) the early sonatas, 5) the miscellaneous pieces and the Diabelli Variations. The initiated will know that even the same concert grand does not stay the same over several months; that exactly the same microphone position -- as if there were a jinx on it -- does not always give the same results; that even technically satisfactory tapes may be distorted beyond recognition in the disc-pressing process. On some of the pressings of the late sonatas the dynamic range was reduced almost to uniformity; moreover, empty grooves of standard length were inserted between the movements, whether or not this suited the context or the composer's instructions, the reason given being that the customers like it that way.

Beethoven's piano works pointed far into the future of piano building. Decades had to pass after his death before there were pianos -- and pianists -- equal to the demands of his Hammerklavier Sonata, Op. 106.

If one tries to play on Beethoven's Érard grand of 1803, which is kept in the instrument collection at the Vienna Kunsthistorische Museum, one thing becomes evident at once: its sound, dynamics and action have surprisingly little in common with the pianos of today. The tone of each single note has a distinct 'onset'; within its intimate confines, it is livelier and more flexible, and also more subject to change while it lasts. The difference in sound between bass, middle and top register is considerable (polyphonic playing!). The treble notes are short-lived and thin, and resist dynamic changes; the treble range is not conducive to cantilenas that want to rise above a gentle piano. Even in the clear and transparent, somewhat twangy bass register, the dynamic span is much narrower than on our instrument. One begins to see the reason for the permanent accompanying piano in the orchestral textures of Beethoven's piano concertos -- even though, admittedly, the orchestral sound of his period cannot have been much like ours. If I had to compare the demands the Érard and the modern Steinway make on the physical power of the player, I would tend to think in terms of those made on a watchmaker and on a removal man!

A few years later, with the pianos of Streicher and Graf, a new, more rounded, more even and neutral sound came into being which, while dynamic scope continued to increase, became the norm throughout the nineteenth century. This sound is more closely related to the piano sound of today than to that of the older Hammerklavier, whose timbre was still derived from that of the harpsichord and clavichord. But by the time this new sound had become established, Beethoven had already composed a large portion of his piano works, and was afflicted by deafness.

We have to resign ourselves to the fact that whenever we hear Beethoven on a present-day instrument, we are listening to a sort of transcription. Anyone still having illusions about that will be disabused by a visit to a collection of old instruments. The modern concert grand, which I naturally used for my recordings, not only has the volume of tone demanded by modern orchestras, concert halls and ears; it also -- and of this I am deeply convinced -- does better justice to most of Beethoven's piano works than the Hammerklavier: its tone is far more colourful, orchestral, and rich in contrast, and these qualities do matter in Beethoven, as can be seen from his orchestral and chamber music. Some of the peculiarities of a Hammerklavier can only be approximated on a modern grand -- for instance the sound of the una corda and even more the whisper of the piano stop. (In the studio, however, finesses of this kind did not always turn out as I wished, either because damping noises obliged me to change my style of playing, or because the technical specifications of the microphone did not permit me to go below a certain dynamic level.)

One must translate other characteristics of the Hammerklavier as best one can. The octave glissandi in the Prestissimo of the Waldstein Sonata, for example, were easier to execute on the older instrument: on the deep, heavy keys of a Steinway they can be brought off only by the use of brute force, which causes them to lose their scurrying pianissimo character. Very conscientious pianists, who cannot bear an untidy note, curb the tempo here and play wrist octaves. The only safe method of preserving the pianissimo character of this section without the help of a piano stop lies in imitating the sliding progress of the glissandi by distributing the passages between the hands, while reducing the bass octaves to their lower part.

The variation works do not conform to the concept of Beethoven, the Olympian. Most of them are unknown even to pianists. Beside the sonatas, many of the variation works appear to be outpourings rather than structures. This is in the nature of the form, which derives from the improvisatory treatment of given material. The attraction (as well as the unevenness) of many variation works stems from the fact that something of the casualness and spontaneity of an improvisation survives in them. The charm of the moment, the lightness, mobility, sharp characterization, the humorous turn are here more important than organic growth. (Admittedly, this does not apply to the masterpieces of the genre: the Diabelli Variations, the Op. 34 and Op. 35 sets, and possibly the problematic C minor Variations.) In the witty, roguish finales we get a glimpse of Beethoven's art of improvisation, which otherwise only manifests itself -- in a different, more passionate vein -- in Op. 77, the Fantasy without basic tonality. Beethoven's at times rather peculiar sense of humour disports itself quite freely here -- as for instance in the delightful 'Kind, willst du ruhig schlafen', my favourite piece in the lighter style, or in 'Venni amore'. In the 7th, 16th, 21st and 22nd variations of 'Venni amore', incidentally, there are distinct anticipations of Brahms, which made it quite obvious that the bearded successor of Beethoven must have known this work, and also 'Das Waldmädchen'. 'Quant'è più bello l'amor contadino' and 'Nel cor più non mi sento' (both after Paisiello) will give unalloyed pleasure to the innocent mind; also the Six Easy Variations on an original theme in G major. The Variations on 'Rule Brittania' are full of bizarre quirks. It is surprising that some of these works made their first appearance in the LP catalogue on this occasion.

What the pianist can learn, and the listener enjoy, in the variation works will be of advantage to both of them when they approach the sonatas. The variation works teach promptness of reaction, exactness and delicacy of characterization, and the ability to regard each variation as having its own separate identity. When compared to the suite, with its well-established formula of movements, the sonata too contained many new personal, private, characteristic elements which must have baffled the eighteenth-century listener. We also learn to be wary of over-dramatization in the sonatas, and begin to see the concept of the heroic Beethoven as a one-sided view representative of the bourgeois nineteenth century.

The miscellaneous piano pieces show us that Beethoven was also a master of the small form, though he rarely turned his attention to it. They are either loosely-gathered and small-scale collections, such as the Bagatelles Op. 33 and Op. 119, and the Écossaises; or they are held together by an inner unity, such as the 'trifles' of Op. 126 -- lyrical, visionary, removed from the realm of the burlesque -- with which Beethoven took his leave of the piano. By themselves stand the spirited, sparkling Polonaise, the hectic Fantasy, and the three Rondos: the two gracefully feminine ones of Op. 51, and the wild, masculine alla zingara work of his early Vienna days.

This last piece has an interesting history. It was published posthumously in 1828 under the title 'Die Wuth über den verlorenen Groschen ausgetobt in einter Kaprize' ('The Rage over the Lost Penny, Vented in a Caprice'); but only in 1832 was the so far unused opus number 129 affixed to it. In contrast to Czerny and Lenz, Hans von Bülow insisted with almost comical emphasis that this Rondo was a late-style work, repudiating any doubt on that point as 'worthy of the Kalmuck Oulibichev' [Translator's note: Alexander Oulibichev (1795-1858), an early Russian biographer of Beethoven.]. The manuscript, discovered by Otto E. Albrecht in 1945, refutes Bülow's (and Riemann's) view: it also contains sketches for works dating from 1795-98. It can be deduced from the state of the manuscript that it served as the basis of the original edition, which was prepared by an unknown hand (Czerny? Schindler?), and certainly not by Beethoven himself. The title current today has been added to the manuscript in different handwriting. In Beethoven's own hand are the superscription 'Alla Ingharese quasi un Capriccio' and the designation 'Leichte Kaprize' on the fly-leaf. The manuscript bears all the marks of a sketchy first draft: uncompleted passages, particularly in the left-hand accompaniment, mistakes in part-writing, and a complete lack of dynamic markings and articulation signs. Unfortunately, it was only after my recording that I came across Erich Hertzmann's thorough investigation of the autograph in the Musical Quarterly, XXXII, 1946, as well as the manuscript itself, so that not all the mistakes of the original edition have been expunged from my performance. An exact text of the piece can be found in the edition I prepared for the Wiener Urtext series.

The beautiful 'Andante favori' and the C minor Allegretto are remnants of Beethoven's work on the Sonatas Op. 53 and Op. 10 No. 1, respectively. The wonderful, well-known 'Für Elise' (or Therese) and another, later Albumblatt in B flat major, a memento to Marie Szymanowska, can hold their own with the best of the Bagatelles, while the Op. 119 and Op. 126 sets look ahead to the Romantic cycles of Schumann, from Papillons to Kreisleriana.

The study of a composer's works appears to me a more profitable pursuit than any pilgrimage to tombs and shrines, or, for that matter, the perusal of a large quantity of critical writing about him. A great deal has been written about Beethoven's sonatas, most of it of negligible value. (On the other hand, despite Tovey's outstanding attempt, I do not yet know of any exhaustive analysis of the Diabelli Variations.) Generally, all one can expect is a little amusement, albeit at the author's expense: thus, a Beethoven biographer from the beginning of this century tells us that the Waldstein Sonata 'had at some time acquired the nickname "Horror", presumably because of the thrusting, agitated figuration and the surprising modulations of its opening which are apt to make one shudder.' The author's shudderings are based on a misunderstanding: the Waldstein Sonata is known in France as 'L'Aurore'.

Among the older Beethoven literature, the commentaries of Czerny, Über den richtigen Vortrag der sämtlichen Beethoven'schen Klavierwerke ('On the proper performance of all Beethoven's works for the piano'), newly edited by Paul Badura-Skoda, are well worth reading; of slightly lesser importance are the writings of Schindler, Ries and Lenz. The interest of Prod'homme's book The Piano Sonatas of Beethoven lies in its inclusion of Beethoven's sketches.

Czerny, taking the Cello Sonata, Op. 69, as an example, describes the repeated striking of two notes connected by a tie, an effect later known as Bebung, which is surely also intended in the Adagio of Op. 106 and in the recitative of Op. 110. The information he gives on the later works is scanty. Yet he does make this comment on the variations of Op. 109: 'The whole movement in the style of Handel and Seb. Bach' -- a remark which startled me only for a moment. It is rather amusing to see how indignant the self-important and unreliable Schindler, Beethoven's first biographer, waxes about Czerny's clumsy, but generally sensible and honest, commentaries. Anyone nowadays venturing to play the first movement of Op. 10 No. 1 in the manner recommended by Schindler would cause some shaking of heads. His suggestion to add two crotchet rests between each phrase in bars 16-21 makes the passage sound rhetorically overblown, while his addition of two fermatas in bar 93 and a caesura before the fp in bar 94 I find downright silly.

Among the more recent books on the sonatas, that of my teacher Edwin Fischer is outstanding; while containing only a fraction of what Fischer had to say about these works, the loving care with which his often quite unobtrusive advice is given makes it more useful than many more exhaustive investigations. Bülow's and Schnabel's editions of the sonatas may, on account of their copious footnotes, jokingly be counted among the Beethoven literature. Both deserve respect as manifestations of strong personalities, and are highly stimulating owing to the temperaments of their authors. Both frequently invite disagreement. Bülow is the first editor to be credited with the attempt to retrace mentally Beethoven's compositional processes; unfortunately, his intellectual method was not equal to his purpose, and he did not pay sufficient attention to the original material. Schnabel, whom I respect as one of the great pianists of his time, was in many ways anti-Bülow: he removed the latter's autocratic 'corrections', but accepted a number of obvious mistakes in the original texts with a kind of pedantic deference. In his choice of reading, I find Schenker generally more convincing than Schnabel, who is said to have been not too happy about his edition in later years. Both Bülow and Schnabel invented highly original fingerings, as did d'Albert, who had a fondness for playing bass notes with the thumb; in his comments, however, he was more sparing of words than his colleagues. At the well-known disputed passage in the Hammerklavier Sonata, before the entry of the first movement's recapitulation, he just says 'A sharp, of course'. As a matter of fact, I play A natural.

Of all the analyses, those by Tovey, Schenker and Ratz (Op. 106) proved more helpful to me than Riemann or Nagel. [Since this article was written, at least three new books merit attention: Jürgen Uhde's Beethovens Klaviermusik (Reclam, 3 volumes), Rudolph Réti's Thematic Patterns in Sonatas of Beethoven (Faber & Faber), and Charles Rosen's The Classical Style (Faber & Faber).]

For a player to study autographs and first prints is more than a hobby; in spite of modern Urtext editions, it is frequently a necessity. When does an Urtext edition deserve to be so called? When, basing itself on all existing original sources, it reproduces the text as the composer might have wished to see it, while at the same time discussing mistakes, omissions and doubtful passages in detailed critical notes, quoting all divergent readings, and substantiating editorial decisions. Heinrich Schenker's exemplary edition of the sonatas and the widely esteemed Henle edition come closest to these requirements, without entirely fulfilling them; the edition by Craxton and Tovey regrettably ignores many of Beethoven's articulation markings, while giving phrasing indications of dubious value. The definitive editorial work is still to be done.

Using the early prints as the point of reference, I myself corrected a large part of the variation works, since at the time of recording no tolerably reliable edition of the second volume was yet in existence. In connection with the recording of the miscellaneous pieces (Bagatelles, Rondos, etc.) I began to prepare an Urtext edition of all those pieces I was including in my gramophone series. Certain important documents, however, did not come to my notice until after the recording sessions were over -- as for instance the autograph of the 'Easy Caprice' and the London first editions of some works, the significance of which was not realized until Alan Tyson's book The Authentic English Editions of Beethoven (Faber & Faber) was published in 1963.

Let me give one or two examples:

The London first edition (The Royal Harmonic Institution) of the Sonata Op. 106 has in bar 116 of the Adagio as first semiquaver in the right hand an F sharp, in contrast to the usual D sharp of the Vienna first edition (Artaria) and all later editions known to me.

This F sharp not only strikes me as stronger and nobler, it also fits better into the melodic line of the second subject: the three-note motive (rising third, falling second) determines its structure up to bar 120.

In other cases I mistakenly relied on the well-known Urtext editions, as the following will illustrate:

The six-times repeated F1 of the pedal point in bars 373-78 of the Fugue in Op. 106 was tacitly provided with ties by Schenker, but these belong only to the overlying trill on B flat1. The logical argument in favour of restriking these notes is furnished by bars 279-80: the sixfold F1 reappears here, this time in rhythmic diminution.

In the Polonaise, Op. 89, we find the following sequence (bars 19-21, also 64-66):

The Henle edition altered, without comment, the bass of the third crotchet of bars 19 and 64 into B flat [In recent reprints the error has been corrected.], thus depriving this pianissimo passage of its special harmonic piquancy. Both these examples are in contradiction to the sources.

I have since changed my mind about the execution of certain details, so that today, in the ninth variation of the Eroica Variations Op. 35, I would play the acciaccaturas in bars 13-17 not before, but together with the left hand. In some cases my reading was inaccurate, or my fancy permitted itself an indefensible variant, as in Op. 28, second movement, bars 72-73. I apologize!

Every generation of musicians is unconsciously influenced by the editions with which it has grown up. My own generation, at least in Central Europe, became accustomed to using editions which respect the text of the composer. Yet necessary though it is to reject the accretions foisted upon the music by the older editors, the restored text is all too easily invested by its users with an autonomous significance which it does not merit. All of us are apt to forget at times that musical notes can only suggest, that expression marks can only supplement and confirm what we must, first and foremost, read from the face of the composition itself.

I should therefore like to propose that the words Werktreue* and Texttreue* be banished from the vocabulary. They have become the feather bed of the academic Classicists. The 'fidelity' referred to here smacks overmuch of 'trust': blind trust, that is, in the self-sufficiency of the letter; trust in the notion that the work will speak for itself as long as the interpreter does not interpose his personality. Let there be no misunderstanding: it is far from my intention to set myself up as the advocate of self-indulgence. The virtuoso who unhesitatingly adapted the music of the past to his own style of playing and composing belongs to a bygone age. Gone are the days when the 'edition', the revision made by a famous virtuoso or teacher, was more important than the original text. That state of affairs, commonly associated with the successors of Liszt, dates back, incidentally, to much earlier times. Carl Czerny -- the teacher of Liszt and pupil of Beethoven -- did not have any scruples about publishing with Diabelli under his own name a Grand Duo brillant à quatre mains, with the minutely engraved subtitle 'arrangé d'après la Sonate de L. van Beethoven, Oeuv. 47'. This is nothing other than a piano duet arrangement of the Kreutzer Sonata!

*[Tranlator's note: Werktreue, commonly used in German musicology, signifies the performer's fidelity to the intentions of the composer, Texttreue his fidelity to the text of a work. Since no simple English translation offers itself, I have decided to use the German terms in the context of this essay..., in the hope that -- pace the author's distrust if not of the words, then of their implications! -- they will be accepted by the English music lover with the kindness he has bestowed on the word Urtext.]

The Romantic era did not yet know a historically-minded style of interpretation. People played everything 'the way they felt', their minds scarcely accessible to arguments of historical propriety. (Editors like Bischoff and Kullak, who aimed at meticulously cleaning the musical text of additions, remained outsiders.) This attitude appeared natural and legitimate until the time when tonality began to disintegrate. In other words, when one could no longer compose or improvise 'modern' cadenzas for Classical piano concertos, the practice of interpreting Classical works in a 'modern' manner also became obsolete. The innocent self-assurance of the virtuoso was gone -- a revolutionary development indeed. In its wake there evolved the editing techniques of the Urtext publication. The investigation of the performing traditions of past ages produced, apart from some misconceptions, a number of genuine insights, which affected the style of Mozart interpretation in particular. The first era of 'historicism' in the short history of public concert-giving had dawned. Its repercussions on the Late Romantic age itself both fascinated and inhibited more than one generation of performers. The loss of self-confidence was often followed by a rigid faith in the letter. People began to play every sort of appoggiatura on the beat, and string quartets would play all four parts equally loudly just because Beethoven had marked them all the same forte or piano. The harshness thus created was considered Classical by many, and still is today in some circles. The dogma according to which every whim of the composer, however unreasonable, must be accepted with reverence, absolved performers from the effort of thinking for themselves. No engraver's error in a first print, no slip of the pen in an autograph was so absurd that it would not be hailed as a bold stroke of genius.

During the same period, the gramophone record established itself. At first a convenient means of preserving the fleeting, unrepeatable impression of a performance, the record, and with it the recording artist, soon laid claim to greater things: all elements of improvisation must stand back in favour of an ideal performance, a definitive rendering divested of any fortuitous aspects. The taking of risks -- for which one needs self-confidence -- lost its attraction and relevance. The image of the machine in its impassive efficiency gained power over many minds; it became an obsession to strive for perfection. In mistrusting their own nature, artists denied themselves access to the nature of music. The usual symptoms of this are that emotions become either completely dried up or wilfullly superimposed. Often, both extremes are to be found in the same person; the vital area between them remains largely unfrequented.

We artists of today have to bear the burden of this paternal heritage, and we feel drawn towards the great ones among our grandfathers. Most younger musicians of all countries, for instance, will readily agree in their admiration for the conductor Furtwängler.

What, then, should the interpreter do? Two things, I believe. He should try to understand the intentions of the composer, and he should seek to give each work the strongest possible effect. Often, but not always, the one will result from the other.

To understand the composer's intentions means to translate them into one's own understanding. Music cannot 'speak for itself'. The notion that an interpreter can simply switch off his personal feelings and instead receive those of the composer 'from above', as it were, belongs to the realm of fable. What the composer actually meant when he put pen to paper can only be unravelled with the help of one's own engaged emotions, one's own senses, one's own intellect, one's own refined ears. Such an attitude is as far removed from sterile 'fidelity' as it is from transcription-mania. To force or to shun the 'personal approach' is equally questionable; where this does not come of itself, any effort is in vain.

The second requirement, that of giving the music the strongest possible effect, can be seen as an attack on the same problem ('What is appropriate to the music?') from a different angle. But let no one imagine that the greatest possible effect can be equated with the noisiest -- or, for that matter, the least noisy -- public acclaim. The crucial distinction is not between, on the one hand, that incessant, extrovert high tension so beloved of naïve listeners, and, on the other, the kind of music-making that fancies itself in the garb of a penitential hair shirt. Those performances that are historically 'most correct' are not always the ones that leave us with the most cherished memories. It would be wrong to modify such memories after the event on Christian Morgenstern's humorous principle that 'what may not be, cannot be'. It is our moral duty to make music in as visionary, moving, mysterious, thoughtful, amusing, graceful a manner as we are able to; but this raises the question 'What is it that will move, shatter, edify or amuse our contemporaries?' There results the paradox that a consummate musical interpretation in which time and occasion seem to have been transcended, in which the shackles of historicism appear to have been broken and thrown off, can only be achieved in concord with our own age. The musical master-work is a power-house of multiple energies. To release those that will strike the noblest, the most elemental resonance in modern man -- it is this task that raises the Urtext interpreter above the status of museum curator. A task, also, that should restore to him some of his lost self-respect.