For no particular reason, I recently relistened to all eleven versions of Robert Schumann’s Carnaval currently in my collection, in chronological order of when they were recorded. It had been a while since I last heard some of them, and some of them I enjoy so much I listen to them regularly. Of course, such concentrated listening also allows one to gain an even greater appreciation of the music. So here goes:

I started the survey with Alfred Cortot’s ancient 1928 recording in the Seth Winner transfer on Pearl. Of all of the pianists I heard, Cortot is the one who most easily conveys the unrestrained romantic nature of the piece. He has his technical short-comings to be sure, but no one else is as effortlessly mercurial as Al. From the start it is a wonderful rendition. One need only listen about fifteen seconds or so to hear the first errant chord, but even with the flubs, his Preambule sets the stage for what’s to come. Pierrot and Arlequin are exquisite, and when Cortot arrives at two of the most important pieces – Eusebius and Florestan – the results are musically spot on. Eusebius is introverted and dreamy, Florestan terse and fiery. When Chopin arrives, it is in a Ballade-esque form, so wonderfully recalling the little Pole’s sound world. The Pause and the beginning of the concluding March of the Band of Brothers are reckless, but my how he enthralls, almost enticing one to go the neighbors’ and grab that TiVo remote from their hands and smack those slovenly zombies upside the head until they understand the value of real art! The ancient sound is expertly redone, and for all of Cortot’s errors, his is a remarkable performance and one to be heard and cherished.

The next stop in the survey is one of the most imposing performances on record: Sergei Rachmaninov’s 1929 recording. From the first notes this is an awesomely controlled but occasionally contrived reading. It becomes more Rach’s Carnaval rather than Schumann’s. The more wistful pieces can devolve into syrupy sentimentality, as is the case with Eusebius. And sometimes Rach uses his power to mischaracterize a piece, as when he makes Repliques sound almost like a lost Mussorgsky piece. But on the upside, of all the versions I heard, his Valse allemande is the only one that sounds like an actual waltz that one could dance to. It’s really remarkable. The Pause is so fast and accurate that it makes poor Cortot sound decidedly lacking in comparison, but the concluding March is too fast and not quite powerful and invigorating enough. All told, this idiosyncratic reading is one to be listened to on rare occasions.

But the next one should be heard often: Leopold Godowsky’s 1929 version stands near the apex of my collection. The first time I heard this I was awed. Here is a version so perfect in mood, tone, and execution that one can’t help but revel in its near-perfection. While he has nearly the same technical ability as Rachmaninov, Godowsky’s version is more fluid and better characterized. When I played it, my fond memories were temporarily brought into question because the Preambule is somewhat weak, with the opening theme not strong enough and the following section not dazzling enough. But it’s smooth sailing from there. Godowsky’s tone is simply glorious: rich, full, colorful, and warm. When he plays Eusebius one is touched by the dreamy lamentations of a tender soul. He can savor the most beautiful, moving passage without ever sounding syrupy. His Florestan, in contrast, is less fiery and more blustery, but still perfectly impetuous. From that point to the end, it is simply glory after glory. Whether playing elegant yet vigorous little butterflies or homages to other great romantics, Godowsky is in his element. And then the March arrives. It is perfectly paced and scaled, and it’s played in an almost breathless fashion at times. All in all, this is exalted music making.

For a change of pace I then went straight to a seemingly unlikely Schumann interpreter: Robert Casadesus. His 1955 recording displays all of those traits that make him one of my favorite pianists: it is crisp, driven, energetic, and elegant. These traits come in very handy in the opener, and they unlock some wonderful secrets in other pieces. Arlequin is jaunty and witty, with just a dash of sarcasm. Eusebius is tasteful, elegant, and restrained, with something of a bittersweet, nostalgic air, as though recalling dreams of youth. Florestan, too, is restrained, with knowingly reduced ardor. Papillons comes off a bit strong, perhaps, but glides along effortlessly. ASCH-SCHA is the best of all of the versions, filled with drive and rigor yet still almost danceable. The concluding March is rousing and full of conviction – Bob apparently saw a need to stand with his brothers – but it lacks that little extra something that the best versions bring. All told, an outstanding version.

It was then time for me to hear the first of two recordings by Michelangeli. His 1957 DG account is beloved by some, no doubt disliked by some, but one thing it certainly is is a marvel of perfect pianism. It is sculpted almost to perfection, if you will. Never is anything out of place or overwrought. The Preambule contains some powerful, exciting chords, and announces that a great piece of music will unfold. Arlequin is puckish but also obviously calculated, but that’s okay, because I love Michelangeli’s delivery. Eusebius starts soft and stays that way, never really firing the imagination as it should, and even Florestan, despite the passionato description never really becomes overtly passionate, but in the context of this performance it doesn’t seem to matter. Papillons is amazingly fleet and perhaps a bit much given the title, but one must marvel at the playing. Chopin is Chopin, though! I mean that in an almost literal sense: no other pianist so ably melds these two great composers together and gives Schumann’s writing that unmistakable Polish air. Pantalon et Columbine just explodes, taking one by surprise, and then it slows to a gentler pace just to explode again! It’s a flawless back and forth not matched by anyone else. The concluding March is martial and imposing, but it doesn’t really make one want to scream out in favor of true art the way Cortot does. All told, it is a superb version, but it doesn’t quite scale the same heights as Cortot or Godowsky. There’s something too calculated and forceful about it. It needs some tempering, some wisdom. Perhaps the older Michelangeli can deliver the goods. We shall see.

The parade of great pianists just continues on to Annie Fischer. I am an avowed devotee of Ms Fischer, finding her unabashedly romantic approach to romantic music pretty much beyond reproach. Yes, she loses concentration here and there, and perhaps she’s not as scrupulous to the letter of the score as other, younger pianists, but, like Cortot, she is always faithful to the soul of the music. And so it is here. Her 1959 EMI recording, in the best sound of the survey up to this point, is a fine one. Her Preambule is a muscular, overtly Teutonic opening, and is almost as impulsive as Cortot, but with better technique. Eusebius is graceful, even tender at times, and Florestan is young and fiery at heart. So far so good. Then comes Chopin. Her Chopin is most decidedly not Chopin-esque, it’s more like young Beethoven. Likewise, her Valse allemande is not very waltz-like, and she ends up roughing it up a bit. But, to compensate, the concluding March is sizzling, especially the very end. Overall, her Carnaval is one to keep and pull off the shelf for a reminder of just how good she is in this repertoire (as this year’s BBC Legends release of other Schumann works by her also ably demonstrates).

The venerable Wilhelm Kempff was next on the list. I’m also an avowed devotee of Kempff, having more recordings by him than any other artist. Now, he was a septuagenarian when he recorded his Schumann for DG, and that means that some allowances must be made for technique from the start. The Preambule shows these shortcomings quite early: the entire piece is stodgy and labored. These traits reappear from time to time. But Arlequin displays some unexpectedly terse playing with remarkably clear fingerwork, so not all is lost. Indeed, Eusebius is almost ideally introspective and dreamy, a true wonder to listen to. Florestan, by way of contrast, displays precious little youthfulness. Papillons is flitting and fleeting enough, but lacks enough substance. Chopin is achingly beautiful, but is ultimately more informed by late Brahms than Chopin. The Valse allemande is really quite poor, being heavy, slow, and undanceable, and the Pause is lumbering. The March is surprisingly stirring and played with utter conviction, but it’s just a tad too slow. A great reading this is not; had only Kempff recorded this twenty years earlier. There are his inimitable grace and tone to listen to, though. One for Kempff fans only.

And now it’s time for a return to Michelangeli for his 1975 EMI recording. Take all of the caveats I mentioned about his 1957 recording and forget them, and just remember the strengths. This is about as close to perfection as I can envision any pianist ever coming. Every note, every dynamic shift, every acceleration or deceleration (often barely perceptible), every use of the pedal has been thoroughly thought through. Nothing is out of place. The immeasurable attention Art lavished on these little pieces produces one of the most beautifully characterized collections one could wish for. Perhaps everything is helped along by the outstanding recording and the absolutely gorgeous tone that is produced: it is warm, never too percussive, and often beautifully bell-like. It is, in a word, flawless. Perhaps some will find that a flaw, hearing this as a bit too mannered and detached. But not me: this is a mesmerizing, unforgettable account.

Not all recordings are so good, of course, and when I returned to Jorg Demus’ recording on Nuova Era I was reminded of that. His cause is not helped by a shallow, brittle recording, but listening through that shows the Preambule to be choppy and mannered, though the playing is crisp and articulate. Eusebius comes off excellently, taken slowly and lovingly with the arpeggios about mid-way through delivered with a delicate, nuanced touch. A weak Florestan indicates that Demus is more in touch with the dreamer-poet than rash young man. (That’s not necessarily a bad thing.) Throughout the work, he seems a bit taxed during more challenging portions, as in Paganini where he seems to be playing at his limit. The concluding March is again choppy, though it is quite forceful, but it doesn’t inspire one enough. Overall, this is clearly not a top-notch performance, nor do I think I’ll be revisiting it too often. Demus does shine in some Schumann, just not here.

My next recording is of another old man, and this one is live. I write, of course, about Shura Cherkassky’s 1988 performance on Decca (and Ermitage). This is a perplexing performance. The Preambule is slow, lumbering, and appears to tax Cherkassky. Pierrot just reinforces that impression. But then in Arlequin he comes out swinging. Perhaps he hadn’t gotten into the piece until then, who knows? Eusebius is wonderful, and played in the Cortot-Godowsky manner, and Florestan is wonderfully impetuous and mercurial. This is some old-school romantic pianism. His Papillons is near perfect: alert and swift, yet never to heavy for the piece. A wonder. Reconnaissance, too, comes off superbly, with the steady left hand underpinning the whimsical right hand. The Pause returns to the troubles of the opening, but his concluding March is grand, almost symphonic in scale, and just a delight, even if it falls short of the best. So, here is a fine set, though ultimately one that must be considered second-tier. If you like old-fashioned, safety-last romantic playing, this may be one to consider, though.

Now for the concluding version: Nelson Freire’s 2002 recording on Decca. When I first got this earlier this year I was not too fond of it, finding it a bit mannered and too focused on speed. With some additional exposure, however, it has grown on me. His Preambule is really quite impressive: after a potent opening theme, he moves to a now dazzlingly swift second section, and that trait continues throughout. Partly as a consequence of his overall approach, Eusebius is not quite up to snuff, any tenderness being all too quickly passed over, but his Florestan is most decidedly a young man in a hurry. It’s quite intriguing and unique. And effective. The next standout is Chopin, here presented as a blend of the early Ballades and middle Nocturnes. The un-waltz-like Valse allemande gives way to a truly virtuoso Paganini – where better to resort to outright virtuoso playing? – before returning the beautiful but undancing waltz theme. The Pause is played as though shot from a cannon, it’s so powerful and fast, and the March is swift, energetic, and rousing. Overall, this recording is a bit undercharacterized with not enough contrast between the competing elements of Schumann’s compositional personalities, but, somewhat ironically given Freire’s age, the work is presented as a young man’s Schumann. It is far more effective than I first thought, and I rather enjoy listening to it.

So, after all of that, it is clear that Michelangeli in his EMI recording is king of the hill for me. But with the early, unabashedly romantic versions by Cortot and Godowsky, and the more muscular, powerful Fischer, and the more elegant and restrained Casadesus out there, it is clear that just one version could never do. Indeed, as I worked my way through the different versions, I was reminded anew of why I collect pieces and why recordings are so valuable: different recordings capture different artists emphasizing different aspects of great art. It’s just not possible for one artist to capture the truth of a piece, because that truth is mutable and changes with time. And it changes with mood. Such a survey also is enjoyable just to hear how musical presentation changes over time, from purposely desynchronized left and right hand playing to a greater emphasis on playing what’s in the score. What is indisputable is that Carnaval is an unqualified masterpiece of the solo piano literature, and deserves – nay demands – to be heard again and again. Go forth and listen!