When EMI made their various sonata and concerto recordings with Claudio Arrau in the 1950s, his reputation in this country was at its zenith; and rightly so, to judge from much that their five-CD Beethoven Edition has to offer. Later, a reaction set in, something that first became apparent in these pages in 1963 as Arrau, now a Philips artist, embarked on his cycle of all 32 Beethoven piano sonatas. It is, I should say at the outset, a historic achievement that has always struck me as magnificently complementing the great rival cycles by Schnabel (EMI) and Kempff (DG). I once described the completed set as ''a journey involving a quality of discipline and selfless involvement at times almost Dostoevskian in its intensity'', and, predictably, coming back to it has been an ordeal, emotionally and intellectually. And here perhaps is the rub. As one distinguished pianist remarked to me recently, we live in an age of easy access and diminished attention spans, which is not perhaps the best environment for a giant like Arrau. At the same time, fashions change. Though SP and Julian Budden were occasionally troubled by aspects of Arrau's Beethoven on record in the mid-1960s, they were generally admiring and respectful. By the late 1980s, though, a CD reissue of Arrau's revered recordings of the Debussy Preludes (revered, that is, by an older generation of critics and rather a lot of fellow artists) was unceremoniously dismissed in these pages as ''leaden-footed'', ''unatmospheric'', ''laboured'', ''stiff'' and ''cumbersome''.
There is, alas, more to being a great pianist than the ability merely to play the piano. Arrau played the piano supremely well but his absorption in the letter of a composer's text and his quest for imaginative oneness with the music's innermost spirit often led to playing that was often more measured, more given to emotional and rhythmic indirection than contemporary taste permits. It is important to remember, though, that such idiosyncrasy was generally reserved by Arrau only for the most challenging and complex music—whether it was a doomed Debussyan landscape or the complex charges of something like the first movement exposition of the Hammerklavier Sonata. Time and again in the cycle of 32 sonatas he is the very model of unaffected good sense. Typically, Arrau doesn't condescend to the two Op. 49 Sonatas, but he doesn't complicate them either. The performances reveal a mature mind and cultured technique that is light years away from a schoolroom run-through; equally, 'interpretation' is entirely in abeyance. You will find this mixture of directness matched by an impeccable keyboard manner in the three sonatas Op. 2 and the two sonatas Op. 14. Arrau is also splendid in the little Op. 79 Sonata, alla tedesca, even though I still prefer Backhaus in the rumbustious finale.
When the Philips recordings of Opp. 110 and 111 first appeared in September 1966, Julian Budden channelled doubts about the occasional comma or momentary ritenuto into a splendid piece of advice. ''The best way to enjoy [Arrau's] records,'' he suggested, ''is to imagine yourself there in the concert hall while he is playing.'' That is shrewdly put, and can be easily demonstrated by turning to Arrau's sublime performance of the two-movement F sharp major Sonata, Op. 54. At the same time, Julian Budden put his finger on another problem that seemed initially to beset this Philips cycle: the quality of the recordings, particularly as they first appeared in some curiously foggy LP pressings. In this instance he found an over-close microphone balance giving rise to some brutal fortissimos that bore no relation to the actual sound of this most powerful but least brutal of pianists. The problem is, Arrau was always difficult to record. Not the least of his accomplishments was a piano sound, uniquely rich and articulate, that would 'carry' at even the quietest dynamic levels into the furthest recesses of a hall. Records convey nothing of this. Worse, with microphones nosing their way into every intimate recess of Arrau's Steinway, the resulting sound could seem burdensome and congealed. This is not always the case. The final disc in this set includes a fabulous performance of the Eroica Variations, gloriously recorded. That dates from 1968. But back in 1962 things got off to a sorry start. Even now there is harshness and some distortion in the finale of the Moonlight Sonata. EMI also give us Arrau in the Moonlight Sonata, a mono recording dating from 1950 which spares us over-miked snorts and sniffs, and reproduces more cleanly. Also, the performance of the finale was even more devilish in 1950, a rare example of Arrau playing with mesmerizing conviction at very high speed. (He could do it if he wanted, though he strongly believed that ''speed is the opposite of passion''.) In a sense, EMI's recordings from the 1950s—a mix of mono and early stereo—are the more pleasing because of their evident limitations. They make no claims to be anything other than a good black-and-white snapshot of an Arrau performance, whereas the less well balanced Philips recordings did originally over-pressure the listener.
That said, much of the Philips cycle is very good, with CD remastering revealing a clarity and spaciousness in most of those recordings that post-date the troubled 1962 sessions which yielded the five sonatas, Nos. 11–15. The Hammerklavier was recorded dangerously early, in September 1963, but the result, on CD, is quite an experience: a torrential performance in full Niagaran spate. One of the qualities that most commanded respect from Arrau's fellow professionals was the uncompromising way in which he investigated a great work and then, equally uncompromisingly stayed with his particular insights and technical solutions. This account of the Hammerklavier is the most orchestral-sounding of all recorded versions; it also takes a remarkably free and penetrating view of the massive exposition, and offers an account of the slow movement that puts Beethoven's music on a par with narrations in Sophoclean tragedy.
One of Arrau's several virtues is that he dodges nothing that Beethoven hurls at him. Occasionally, it is true, he finds things to worry over which lesser mortals pass urbanely by. But collectors who wish to hear domesticated Beethoven have plenty of other pianists to whom they can turn. I wouldn't put Kempff into that category, but as I remarked in my review of his set last March, on the matter of repeats, Kempff can be worryingly cavalier. He even goes so far as to omit the exposition repeat of the Hammerklavier. Beethoven agonized long and hard over the question of repeats and Arrau's respect for his decisions is total, right down to taking the repeat of the second half of the first movement of Op. 2 No. 2. Here the playing is so full of vitality and relish that few will complain. And it is interesting how Arrau goes from there, via a profound and slow reading of the Larghetto and a skittish Allegretto, to a wonderfully wide-ranging view of a finale that shows the young Beethoven in tough and visionary mood, the ladder-like upward arpeggios decisively unfurled. In the end, Arrau takes nearly 30 minutes over the sonata, where Schnabel and Kempff, with only one first movement repeat and more rapid readings of slow movement and finale, see it through in just over 20. Perhaps that is right for an early sonata; and yet, what do we mean by 'early' in the context of a genius of Beethoven's magnitude? With Arrau, Op. 2 No. 2 is simply (or not so simply) another towering Beethoven masterpiece.
Occasionally, Arrau's breadth borders on eccentricity. His account of the quick movements of Op. 2 No. 3 is exemplary. But the Adagio is gruelling: 10'13'' (cf. Schnabel 7'41'', Kempff 7'01'') taxes even my attention span. On the other hand, I don't see the justice of criticizing Arrau for his leisurely way with the finale of the E minor Sonata, Op. 90, where Kempff (surprisingly) is even slower. Schnabel's quickish tempo may be the theoretical ideal. But it is precisely because of Arrau's leisure, and his occasional hesitancy at critical junctures, that we become absorbed in the eventful progress of Beethoven's imaginings. I would even defend Arrau's astonishingly bleak treatment of the Adagio grazioso of the G major Sonata, Op. 31 No. 1. This is usually seen as a parody of rococo vapidity. But when Beethoven wanted to mock musical mindlessness he was invariably simpler in his methods than here. Arrau invents nothing. But he does take us into a strange fantasy world, a kind of rococo Hell, with the left-hand bass trills near the end beating like the wings of a flock of birds of ill-omen.
Nowadays critics often prefer artists' earlier recordings to their later ones: another manifestation, I suspect, of the mindless worship of youth that has dominated our thinking since the 1960s. Certainly, it would be tempting to see EMI's wonderful selection of Arrau's Beethoven sonata recordings of the 1950s as a set of cautionary criteria by which to measure the Philips remakes. There is, for example, a tremendously alert and winning 1947 account of Op. 10 No. 3 (much admired at the time by Alec Robertson). The surfaces sizzle a great deal, but the sound is excellent for its age. Yet the Philips is appreciably broader in tempo (the sonata has no slow movement as such) but with glorious playing and, on this occasion, Rolls-Royce Philips sound, it would be perverse to prefer the earlier disc. On balance, I prefer Arrau's famous EMI account of the Waldstein to the Philips remake. Philips's engineering is superior, especially in the finale's huge climaxes; but, intentionally or not, there is a touch of mistiness and hesitancy about the exposition's first statement in the first movement. This is a pity. The tempo is so finely judged (Arrau despaired here of speed-merchants pecking at the notes ''like chickens in the dust'') and Arrau's reading is so fine-grained, with the crucial grace notes sounded very precisely on the beat as infinitely suggestive appoggiaturas. Both versions, though, give you a dramatic reading that is scrupulous to the letter of the score. Arrau's realization of Beethoven's pedal marks in the finale was always a model of imaginative technical control.
The EMI set also includes a terrific performance of the Appassionata. This is everything one could hope for: dark-coloured and concentrated but also profoundly lyrical and fiercely alive in every fibre of its troubled being. That said, the Philips performance is even more awesome in its outpourings. Daniel Barenboim has made the point that though Arrau often adopted unusually broad tempos in later years he had the genius and control required to fill them. This is certainly the case here. The EMI recording of Op. 101 is another classic of the gramophone—the whole sonata, not just Arrau's playing of the Vivace alla Marcia that has long been the model of how this difficult movement should be tackled. The same could be said of Op. 54 (on EMI and Philips). EMI's Les adieux is very fine though Arrau's forceful tone is once or twice rendered ugly by overawed microphones.
It is frustrating that EMI have lumped their five-disc treasury into a single box. (And what a clumsy and frangible box it is!) I would love to say to younger collectors: ''hasten off this minute to beg, borrow, or steal Arrau's EMI disc of the sonatas Opp. 101, 110 and 111—for in a lifetime's collecting you will never put a worthier disc on your shelves''. Never has the first movement of Op. 111 sounded blacker or more torrential than here (and all achieved as Beethoven wrote the music down, with no faking with two hands where the score implies just one). True, I would like a little more air and light in the opening paragraphs of Op. 110 where Beethoven's writing suggests the free play of a speculative mind; but the playing of the sonata's Arioso dolente (even freer in the Philips set) is lofty and moving in a way that beggars description.
The drawback of the EMI set is the concerto cycle with Galliera. When the recording of Concerto No. 3 first appeared, Trevor Harvey was moved to ask, ''Who chooses conductors?'' and William Mann (who had produced some of the sonata sessions for EMI) was almost equally unhappy with the Emperor. Galliera was a competent musician (he is perhaps best known for conducting the Callas recording of Il barbiere di Siviglia). He ably supports Arrau at several critical moments. Most notable are the various transfiguring transitions (in Concerto No. 4, for example) from cadenza to coda. Elsewhere, the accompaniments can seem dispirited and inconsequential, like the shadow of an old A road running alongside the black swathe of a new motorway. In the Emperor, Arrau not only plays the solo part with imperial strength and magnificently variegated colours, he is also the provider of most of the musical traction. Hans Keller liked to dub this Beethoven's Concerto-Symphony; and in Arrau's performance you are certainly aware of the soloist as symphonist. That said, Arrau's later recording with Davis and the Staatskapelle Dresden on Philips is much to be preferred.
EMI's recordings of the concertos are also technically uneven. Curiously, Arrau seems to have had no settled producer or balance engineer at EMI (a dozen names appear in the credits, plus our old friend Anon). In the earlier concertos, Arrau's playing is enormously cultured. For my own taste there is too much equanimity, and too little humour. I miss the uproariousness and razzmatazz that Beethoven, working within classical forms, brings to these early works. Yet Deryck Cooke, reviewing the EMI recording of Concerto No. 2 in November 1960, found ''great classical purity, virile warmth, and depth of feeling... [Arrau's] wonderfully affectionate handling of the slow movement, particularly the breathtaking lead back to the main theme, is one of the most perfect pieces of piano-playing I have ever heard''. With comments like that, who is to say that this entire five-CD set isn't worth investigating?
In due course, I hope that EMI will put some of this set out separately, and that Philips will also retain the separate CD volumes of the sonatas in the catalogues. Vol. 3, in particular, is glorious. It includes the Appassionata through to Op. 111, plus magnificent performances of the early sets of variations. Even if in the sonatas as a whole you are temperamentally drawn to Kempff or Schnabel, there is a loftiness of vision and a magnificence of address about Arrau's playing that puts him in a class of his own.'