Description
One of the great variation sets of the Romantic period, alongside Liszt’s WKSZ variations + BACH Fantasie & Fugue. For some reason Liszt has never been recognised as one of the most important variation writers, and the Totentanz in particular tends to be viewed primarily as a showpiece. There is a certain truth to this—everything attempted here is so immediately convincing you don’t sense any craft behind it. But as with lots of Liszt’s big works, even a little examination shows you how much there is going on.
For a start: the sheer fecundity of this! The score (weirdly) indicates only five variations, but a conservative count shows at least 21. Even this vastly understates things, because Liszt often includes variations within a variation, often by deploying a completely distinct texture or character for the dies irae’s second phrase. And within these 21+ variations, the work runs the gamut of pianistic/compositional techniques—from toccata exuberance to modal counterpoint,
lyrical diatonicism to diminished-harmony wildness and even a Locrian double-flat 7(!) scale.
Then there’s structure. The Totentanz is structured very tightly (as tends to be the case with all the works Liszt extensively revised) and feels a lot shorter than its 15 minutes. Loosely, this is a cyclic sonata-allegro. An introduction, then a set of seven variations with a slow underlying tempo which are fairly directly connected to the theme. This is followed by a more complex and developmental set of six faster variations, which have their boundaries blurred by aggressive motivic play. A new variant of the dies irae (using the La Folia harmony!) is then introduced with a more stable, diatonic character, and we get six variations on this in rapid succession—essentially establishing a zone of stability after the tension of the “development”. This “La Folia” section finally yields to a huge cyclic coda in which the two of the defining elements of the first section—diminished harmony and glissandi—are reintroduced.
On a bar-by-bar level, there’s some very neat things happening. Suffice it to say this work is dense with motivic play; the first four notes of the dies irae (the “Incipit” motif) are often combined in different variants, so that it’s not out of place to see the motif being played at three of even four levels of diminution/augmentation at once (or in very close proximity). (See the transition at 8:30 for an example.)
Last, this work showcases two of Liszt’s greatest natural gifts as a composer, gifts whose manifestations most casual listeners immediately detect but which are often dismissed as surface-level features (as if that somehow diminishes them) in formal analysis. The first is Liszt’s facility for piano+orchestra textures—at this point he was (by some distance) the greatest conjurer of such sounds that had ever lived, and it shows here. At many points, the writing attains a timbral precision that which only Ravel and Saint-Saens later would match—Var. 17, for instance, or the combination of pizzicato strings/bassoon/piano glissandi in the coda (14:56). These are extraordinarily effective and very nonstandard textures, the kind of things that only spring out of a great musical imagination. The second is Liszt’s inerrant sense of paragraph-scale drama—you see it in the stark opening textures, but also in masterful transitions that sustain tension over multiple variations. See how the latter half of Var.7 (7:05) dramatically sets up the fughetta (Var. 8), which intensifies until it plunges gloriously into Var. 9, which in turn introduces a long transition into the ecstatic apocalypse of Var. 10.
It’s a testament to how well put-together the Totentanz is that it’s more or less interpretation-proof. And interpretative divergences abound. To pick just three spots: (1) The opening. Is the piano forward in the texture, or just there to add a metallic brutality to the bottom of the orchestral timbre? Are the chords a garish staccato? Or are they pedalled into a mash of black sound? (2) The octave hit that starts the transition into Var. 8. Dull ominous whump, or nuclear bass drop? (3) Var. 10. When you reach it, do you slow into a grandioso, or plow the rhythmic throughline to a frenzy?
00:00 – Oravecz
15:38 – Freire
29:57 – Watts
44:59 – Masleev