musical thoughts

The Ralph Towner composition Icarus was the subject of analysis in a discussion about whether melody or harmony was more important in creating tension and release. The precise original question is lost (I’ll post it if I run across it), but this is still a detailed analysis of an interesting, well-composed piece.

I consider Icarus an incredibly effective tune in terms of its manipulations along the tension-release axis. This tune works. So maybe through analysis we can ascertain whether that’s more driven by melody or by harmony. I picked this because it was strong, but simple enough anyone ought to be able to follow the analysis.

Here’s the first time through the form, so you can hear. This excerpt is expressly for educational purposes, please don’t ask to send you the file for the whole track.

Here’s a roadmap that shows the changes, omitting the melody entirely. Spellings are from the chart I have, but may not depict the functions ideally. Key signature on the chart is G.

|| G (9) | | G7sus | |
| G7sus | | Cmi9/G | | | |
| G (9) | | G7sus | |
| G7sus | | G (9) | | G7sus | G7 ||
|| Cmaj7#11| | | |
|Fmaj9 | | | |
| D(9) | | | ||
|| Bmi9/F# | | | |
| G7sus | | | |
| Bmi9/F# | | | |
| G7sus | | | ||
|| C (9) | | Gmi9/C | Gmi9/Bb | EbMaj7#11 | |
|F#Maj7/G | | | ||

Four sections, ABCD

A section is two ten-bar phrases. It’s drawn with the last four bars of each as first and second endings on the chart I have. So the first phrase lands on a chord you might want to consider as the first inversion of an Eb major 7th. It’s not associated to the key signature at all, so it’s probably more useful to look at the whole A section as a pedal point over G. the first three chords could be seen as G, F, and Eb major triads over the pedal. the last four bars of A set up the modulation to C major in the B section.

B section is three four bar phrases, each consisting of a single chord. The circle of fifths motion we established with the cadence into B section continues as C resolves to F. The D chord is kind of a surprise. If we relate it to the original key center (G) it’s a dominant, but I’ll argue it’s really not functioning as one here.

C section is constructed in parallel: two eight-bar phrases with the same progression. You could spell the Bmi9/F# as Dmaj7 in first inversion. there are no common voices between subsequent chords, making the movement feel more like a parallel block chord than like a cadential formula. The repetition builds tension, which resolves via an authentic cadence to C major at the double bar that marks the D section.

D section opens up convincingly in C major, which really opens up doubt as to whether this tune is in G as the key signature indicates. An authentic cadence in G never does occur, but it occurs twice to C. D is a ten-bar phrase that initially seems to imitate A section in the key of C. The first chord movement can be viewed as a major triad moving downward stepwise ovcer the pedal, like in A. But then the root moves beneath it in the fourth bar of D, and then again in the fifth, landing us in an Eb lydian tonality (which got implied in the a section, but the G pedal point kept us anchored). Note that the change Gmi9/Bb -> EbMaj7#11 can be accomplished with just the root motion–all voices are common. And then the final chord of the tune, which deserves a special discussion.

F#Maj7/G, with an A in the melody (!!!) is not exactly where you expected this thing to wind up. On the other hand, consider what happened to Icarus in the myth–he plummeted to earth and suffered the effects of rapid deceleration. This progression ends with a crash rather than a cadence. But it achieves the purpose of the tune (play this progression if you doubt me).

So the principal harmonic devices in play in this composition are:
+ dominant prolongation–you can view A as entirely that
+ parallel motion over a pedal
+ repetition of phrases
+ restatement of material in a modulated key (D restates A up a fourth)

Looking at the melody, the programmatic nature of this piece gets revealed fully. Refer to the complete lead sheet:

The melody has a compass of two octaves, which is pretty wide. The first note is the lowest, the highest is in the seventh bar of B section. Almost every pitch change is on an upbeat. the contour of the melody in A section is skips up, steps down mostly. The octave leap in bar one of B section is especially dramatic, because it is the highest pitch encoutered at that point, and the prior pitch was prolonged nearly four bars, building tension that gets released in the octave leap. We descend from the new high point along a pentatonic scale in the first phrase, and begin the upward contour with leaps in the second, climaxing with that peak in bar 7, and remaining on A for bars 9-12 of B section.

C section melody opens with a long note that is outside the key signature (no matter which one you consider correct, C# isn’t part of it). The climb in the second phrase sets up the modulation to C major in the D section. The parallelism of C section melody creates tension that underscores the tension that comes from the minor tonality.

D section melody “parts the clouds” of that excursion into minor definitively, by restating the A section melody up a perfect fourth. But we descend to the starting note by the fifth bar, unlike the A melody, and the harmonic crash follows, with the non-chord tone A in the melody over the already amazingly dissonant final chord reinforcing the “anti-cadence” that ends the piece.

+++

Now to interpret this craziness in the context of the original question: I’ll argue that the melody and in particular its shape, create the major tension/release interplay in this piece. Icarus comes up, he glides down, he goes higher, he soars, in the end he gets to close to the sun, and then he crashes and burns. Ouch. As kinesthetics go, this melody is as good as you’ll find. Now for sure the harmony underscores this (is that a tautology? perhaps). But there’s the kicker for me–if you isolate the chord progression like we did first, it doesn’t really make people react the same visceral way as the leaping melody of B section does. On the other hand, it’s damn tough to do a “crash” melodically. For that we need the vertical dissonance. And boy, do we get it.