Harmonic Analysis of Three Arias in John Adams’s Nixon in China (Part 2)

            As Timothy Johnson demonstrates in his book, Adams avoids the traditional functional harmony in favour of chord progressions than can be approached with a Neo-Riemannian harmonic language. He uses chords, both triads and sevenths chords, to represent the emotions of the characters and the evolving attitude of the characters towards each other. To avoid discussing too much of the plot, I will focus solely on the music, and not the text. Adams uses non-functional chords in such progressions that it mimics functional harmony. In certain sections some chords can be called the functional tonic and dominant, while in others, chords serve as different levels of “terraces”, either stepping up tension or releasing it. As with all theories, there are a few holes in Johnson’s analysis of Nixon in China, some of which will be addressed to the best of my ability. I do not believe that Adams composed with this “system” in mind, mostly because it hasn’t been invented yet when he wrote the opera; the theory attempts to fit music, just as grammar attempts to fit language. Some things just do not fit neatly within the Neo-Riemannian context.

Click here to view the Score and Flowchart

Notes for “News has a kind of mystery”

            The first aria, Nixon’s aria “News has a kind of mystery” starts on an Ab major triad. For ten measures (mm. 374 – 386), this chord remains before changing to an F minor chord, a relative transformation. Since this is a basic transformation, I consider is a “non-tension” movement. In the Johnson book, he talks about how the Ab and the Fm are from separate hexatonic systems, but I think that the systems are good for categorizing the different transformations, just not so good at analyzing the chords that are a part of the systems. The “set up” not only makes little sense contextually, but also, more importantly, make little sense aurally. The sequence sounds, in a functional way, like I – vi (and to iii when it moves to Cm later). These two chords are all basic transformations from Ab major, and therefore should considered mere colourations of the “tonic” or what I’m going to call “home” triad.

            In mm. 409, the Ab triad goes through a leading tone and then a parallel transformation, a complex transformation, into C major. This, followed by a crescendo, sounds to me like a real chord change, into a new tonic, effectively making the Ab major triad the tension chord resolving into the C major. The C major then goes through another LP transformation into E major, and C becomes the new tonic before finally going to the new home centre, F minor and Db major. From E major to F minor, the chord goes through a SLIDE transformation, keeping only one note constant. This always sounds like a modulation to me. The system of keeping a home centre around two or three chords continues in this aria, and modulations or tension building occurs on compound transformations.

            Section F (starting mm. 542) has a curious section that uses a new technique to achieve tension. Bichord dissonance set in contrasting registers. I can only describe this section as a flowing tension and release, as the chords are connected by many complex transformations. This does not fit with the Neo-Riemannian understanding of the rest of the aria, and I believe that is why Johnson did not include this section in his analysis. There is another section like this one at measure 616, which Johnson just wrote “transitional section” and moved on. The bichords Abm/Db into Bbm/Gb are true non-functional chords, that only adds colour to the melodic line, which stays solely in Db major. The chord change only then occurs in mm. 553, at the F#m chord. From that point onwards, the Fm chord at mm. 560 only serves to reposition the melody. Then the C#m add major 7 chord melts into Ab+ and Ab, which is the home of the next section. I believe Adams wrote this section to colour the existing melody. The F#m at mm. 553 heightens the tension before falling back to “home” at the Fm chord. Then, the C#m add major 7 is the real climax of this section, falling twice, first halfway to the Ab+, losing the bass note, and then finally back to Ab major. Adams uses tension as levels, not tension and release, which will also appear in the next aria.

Notes for “I am the wife of Mao Tse-tung”

            “I am the wife of Mao Tse-tung” uses the dominant and subdominant transformations. A dominant transformation changes a minor chord into its “tonic” as if the chord was the dominant, for example, Bbm into Eb. The subdominant transformation does the opposite.

            Although we have not seen 7^th chords yet, in this aria, they serve as an added layer of suspension, usually played by a separate instrument than the rest of the chord. The A section of this aria has a Neo-Riemannian section, where the voice leading makes conventional sense. The Bb to Bbm to Eb7 to Ebm goes though parallel transformations at the major to minor switch, aurally sounding like a depression to scare you into the dominant which is actually more of a IV sound to me. The pull is still from the Eb (IV) to the Bb (I) and not, as you would probably have guessed a V to a I, Bb to Eb. The basic transformation of the major to minor, is still only a colour change.

By the end of the aria, during the instrumental coda (mm. 961-end), the chords D7, E7 and the home chord, Bb are used to represent different levels of tension. The D7 and E7, both only sharing one note with Bb serve as tension towards the Bb home chord. Since E7 also has a tritone relationship with Bb, it sounds more jarring in relation to D7. In addition, the way the bass always starts on the 1 of the chords during this section makes the D7 sound like it wants to move to the E7, and in turn, to Bb. Therefore, it can be said the D7 serves as a subdominant function chord, and the E7 and Bb serves as dominant and tonic.

A quirk of Adams in this aria that we haven’t discussed yet is at mm. 832, the dominant sound of the Bb chord, with the E7 chords. The tritone distance makes the E7s sound very jarring, and it can be seen as a secondary dominant to the “resolution”: a Gb chord at mm. 852. This isn’t a V-I though, Adams still steps down to it, through a P and L transformation, after the high note. This climax, you would think, needs a climatic harmony to compliment it., but with the transformations, Adams “side steps” your ear, and modulates into a new “home”. This technique of stepping down into a resolution happens very often throughout the entire, and often at traditional cadential points.

Notes for “I am old and cannot sleep forever”

            I am going to diverge once again from the Johnson. He analyses this aria from the melodic solo line and then moves to describe the function of the bichords and how they serve to harmonise the very dissonant melody. I do not believe that the bichords serve a function in this aria. There is a background dissonance from the lower half of the bichord, that being the chord on the bottom, the lower range, serves no functional purpose. When listening to it, all the “dissonance” that Johnson has analysed sounds only like noise, and not additional notes of the chord. In addition, the bichords share many of the same notes as the main chord, and it can be viewed as extensions. Basically, the bichords colour the melody. Therefore, we can analyse the A section as three distinct colours, without tension between the three. E minor, G# minor and C minor are used to propel the melodic line along, but the shared common tone doesn’t allow for, at least from what I can hear, much tension and release.

            The C section, at the tail end of the opera, uses Eb+/Cm chords, or just a C minor major 7^th chord as the “home” chord. And honestly, I cannot figure the rest of it out. From mm. 907 to the end, just the last couple of bars, I really don’t know what Adams was trying to achieve. We know that the opera is ending, and that Chou En-Lai is struggling with his insomnia and his legacy. The muddy mess of colour and chord that populate the final few bars of this opera sound just like that. No function, no purpose, just… noise? It’s still a mystery.

Additional Notes

            For the flowchart, I decided to just mark the flow of the chords out with the form and sections marked in, since there’s very little cadential action in these arias. In addition, I did not use roman numeral analysis for these arias since that would make very little sense harmonically, due to the whole lack of traditional functional harmony.

            There are still some sections in the flow chart that doesn’t make functional sense. The A prime section in the capo of “News has a kind of mystery” where the E half diminished chord moves to the F# minor then to the D of the B prime section makes little sense to me. I labeled it “transitional period” but that’s just to sidestep writing nothing there. The quartal chord throws me off. And so does the 7^th chord since I’ve been ignoring the 7^th chords in general. I think quartal theory and a deeper dive into the newer papers on 7^th chord in a Neo-Riemannian context will be helpful to understand Adams’s compositional technique there. Of course, the very last section of the opera is still foggy to me. And then there’s the transformations. I have found that basic transformations sound little like real functional chord changes, but then then question arises on what counts as chord function. I have always labeled a chord as the “home” or tonic, but that’s just the chord that begins the aria and therefore, what I start analysing on. Then, perhaps, just like Johnson’s theory, this one is also kind of wrong and kind of wonky at certain areas. Things don’t always fall in place in a functional context, like I thought it would. And unlike a scientific paper, I’m not going to try to make the data fit my erroneous hypotheses.