I always had the feeling that Scriabin's early experiments with chromaticism came from the Chopin line, rather than from Wagner. His mature works are unique enough that they sound only like Scriabin and have little in common with Debussy, who also came from the Chopin line. However, there are signs that both men were singed by their exposure to Wagner’s fire. A recent investigation of Poem of Ecstasy was prompted by my hunch that the end of the piece is an homage, or parody, of the opening measures of Tristanund Isolde.
What I always find most interesting about Wagner scholarship is that so much theoretical genius has been spent on the analysis of the two measures that contain the Tristan Chord and its resolution, but almost no one has been able to successfully explain the first three notes (A-F-E). Wagner is playing a joke on us because the piece begins with an unusual tonic (?) pickup that leaps up to F that behaves much like a very, very long appoggiatura resolving back to E. Since there is no C in the neighborhood we cannot know for sure that Am is his intended key. In fact, these three notes work better in D minor. The Fø chord that follows changes the history of tonal theory forever and things only get more complicated from there.
At the end of the Poem Scriabin plays a joke on us by putting those three notes together in an Fmaj7 chord over a tonic pedal (C). The kicker is that his arrangement of that chord has the A (the Tristan pickup) as the next note up from the pedal. The first significant accidental in Tristan is the D# following the E and that is the two-note pair that Scriabin repeats six times, yearningly, just prior to the ultimate cadence. In Tristan the G# that sits on top of the D# is the mystery nonharmonic tone that follows F-E and resolves by semitone to the A and forms the significant French+6 chord (F7b5). In Tristan our ear follows the melodic line from F to E to G#. Scriabin parodies this 1-0-4 motive by moving from E to D# to A (1-0-6); followed by A moving to G# (the retrograde of Tristan's G#-A) to C (1-0-4); and, finally, C-B-E (1-0-5). Intriguingly, this melodic activity spells out an A minor triad (the implied opening key of Tristan) while the key at the final cadence is actually C major.
In Scriabin the penultimate sonority (chord?) is formed by the C resolving downward to B--the inversion of Tristan’s G#-A. In both cases the semitone motion results in the same B7b5 (F7b5).
Some will say that this is all negated by the tonic pedal on which the whole thing sits and is, therefore, nothing like Tristan. May I suggest that Wagner’s final joke on us in the Tristan Prelude is the appearance of the Tristan Chord over a dominant (!) pedal in measures 102–103 (G7b9). At the beginning, the Tristan Chord was a secondary dominant 7th in A minor and, later, it reappears as a dominant in C. Scriabin’s pedal tone is, in my opinion, inspired by Wagner’s. Just to be different his Tristan Chord is an altered common-tone dominant 7th (IVdom7b5) that gives us a very barber shop style final cadence. Whatever you do, do not underestimate the importance of the penultimate B in the melody. It is not just an escape tone. Its dissonant relation to the pedal is significant and is, after all, the leading tone (!). Its significance is heightened by the fact that we have had to wait forty-eight measures for the note B to finally show up as the penultimate melodic tone above the pedal. When the final leading tone in a piece of music turns out to be a flatted fifth you know it is time to say farewell to tonality as we know it. The late Romantics owe much of their tritone obsession to crazy Dick.
If you don’t believe in magic, don’t analyze Wagner. If you cannot accept that composers are delightfully crazy to begin with and often like to play note games and mess with our heads, don’t go into music analysis.
If you are still with me it may be worth your while to stay tuned for just another minute. I was bothered by something that seemed to be missing in my comparison of Tristan and Poem and now I've got it. In order to better understand Scriabin’s harmonic intentions I needed to consider the last two minutes of the piece, not just the last couple of pages.
There is major parallel that needs mentioning. The last 53 measures of Poem sit on a tonic pedal, but the first chord over the pedal is the dominant 9th (V9), not the tonic. After eight measures this dominant moves to the subdominant just the way the Liebestod shocks us with its long dominant that resolves magnificently to the subdominant. It is even preceded by a 0-1-2-3 (F# – G – G# – A) motive from mm. 2-3 of Tristan.
In Tristan the first Tristan Chord is a secondary dominant of the dominant, the final one in the prelude is a dominant of the relative major (C), and....the last one in the opera is sitting on a tonic pedal, just as in Poem, and functions as the dominant of the minor subdominant. Tristan ends with an altered plagal cadence and so does Poem. To me, the penultimate quarter note B in Poem is highly reminiscent of the penultimate quarter note C# in Tristan.
From all these similarities I can only surmise that Isolde’s supreme bliss must have haunted Scriabin and prompted him to create his own musical waves in which to drown.