Note
Please see the Preface to this Series to understand the goals of putting up this unpublished work and the general apologies for not citing a more up-to-date bibliography.
This paper is a (mostly unedited) seminar paper presented to Reinhold Brinkmann's seminar on twentieth-century opera (Fall 1999). Thus no bibliography or citations post 1999 are included. It was also written at a time when many musicologists could have not known the basics of Einstein; now this view seems a little obsolete. The only changes (beyond fixing of typos) in this version are the YouTube clips that have been added where easily found.
It Could Be Very Fresh:
Structure, Repetition, and Reception Einstein on the Beach
(1999; part 2)
Glass's Analysis
Glass has chosen to base much of his analysis of the work on the harmonic features of the music. The introduction of harmonic shifts within sections, begun with Music in Twelve Parts (1971-4) and continued in Another Look at Harmony (1975) which Einstein grew out of was for Glass a significant change in his musical style in the years preceding the opera. It is understandable that since most common tools for analyzing Western music rely on harmonic structure he would employ these methods in looking at his own now harmonically shifting works.13 However, a look at Glass’s harmonic analysis of a single motive, five, reveals how little roman numeral analysis of his motivic idea tells us about the work.
13 Other conventional analytical tools would present other problems; Schenkerian and other voice-leading analysis techniques, even if they were available to a composer who received his musical training in the 1950s, can readily be seen to hold little promise for understanding this work.
If as Glass has said “process here is the subject rather than the source of the music” and “the noticing of process itself becomes exhilarating,”14 then an analysis which focuses on one iteration of a process is missing some of the most salient features of the piece. Glass has pointed out that the motive, f-minor, D-major, A-major, B-dominant7, E-major can be heard as i-VI-IV♭ in f with IV♭ being heard as a pivot to IV-V-I in E. He has further stated that since the motive ends a half-step below where it began it “provides the leading tone for the original i (f). As it is a formula which invites repetition, it is particularly suited to my kind of musical thinking.”15
14 Robert Wilson and Philip Glass, Einstein on the Beach, edited by Vicky Alliata, (New York : EOS Enterprises, 1976(?)), [n.p., center of book]. The quotations sound more like Reich than what we are used to hearing from Glass in their emphasis on the perception of process.
15 Philip Glass, “Einstein on the Beach,” essay printed in liner notes to recordings of Einstein on the Beach and in Music by Philip Glass. Part 2, paragraph 6.
While he is correct in stating the five chords can be heard as a modulation from f minor to E major, it is certainly questionable whether any listener will hear it as such, or especially whether the twentieth repetition of the cell will produce such an effect on the listener.16 The ability to hear this passage tonally is particularly hampered by the voice-leading from the fifth chord to the first chord of the repetition. While the other four transitions between chords followed traditional four-part voice-leading rules,17 the motion from E major to f minor contains three major “errors”: There are parallel octaves between the “alto” and “bass” voices, parallel fifths between the “tenor” and “bass”, and a doubled leading tone. After hearing this non-common practice transition, it is unlikely that the listener will perceive further repetitions tonally; the use of IV♭ as a pivot chord was already a stretch to hear the first time.
16The sections of five in “Train 1” have 39 repetitions each. “Knee 2” has 44 repetitions of five (22 + 22). The 158 repetitions in “Spaceship” dwarf any other section of five in the opera. I will be using the term “cell” to refer to a specific instance of a motive in which the motive may be rhythmically or melodically altered.
17 This statement notes but takes exception to the doubling of the bass a fifth higher which, everywhere except in the score, is heard as an acoustical effect of the electric organs and not an independent voice.
If we do not hear this progression as a tonal modulation (at least after the first presentation) it makes sense to ask whether Glass’s choice of chords has any bearing on our perception of the music. I will argue that does using two alternative versions as counter-examples:
Purely tonal variation
Non-tonal variation
The version above is a strictly tonal version of five, grounded in E major. The lower version is an
atonal rendition of the theme, not cadencing in any key. Both versions
preserve the general bass contour of five, use triadic harmony, and retain three of the chords
of five.18 Yet
neither alternative is satisfying under repetition the way five can be heard to be. In the tonal
version the listener becomes frustrated because there is a strong unfulfilled
expectation that the music will “go somewhere” harmonically yet there is a
feeling of sameness because the music does not—it is simply E major followed
by E major. The atonal version distances the listener for the opposite
reasons. Without any cadential formula there is no harmonic expectation
created and little (harmonic) reason to continue concentrating on the music.
It is the version in Einstein which balances these listening concerns.
The final three chords (IV-V7-I) form the most efficient
establishment of tonal center possible while the augmented triad formed by
the first three is an effective way of eliminating a possible key—the only
triad not able to be constructed from diatonic scales.19 The
listener tries to make tonal sense of the progression only to realize in the
next repetition that this is futile. Later, the strength of the
IV-V7-I cadence invites him or her to try again, repeating the
process many times throughout the piece.
18 With the hindsight of knowing his future output, one might wonder if the atonal version is actually a plausible alternative Glass could have written. It should be remembered, however, that in 1976 the last major work by Glass was Music in Twelve Parts which ended with a complete twelve-tone row in the bass.
19 Another piece from the 1970s which achieves its harmonic interest through a contrast between two contrasting harmonic areas within a repetition is Reich’s Four Organs (1970). An example of a piece which uses harmonic material related to the “atonal” theme is Louis Andriessen’s Hoketus (1977), sections A-D. As in the alternative Einstein version, the two and later four non-tonally directional chords of Hoketus cause the listener not to derive interest from the pitches played within a section. In Andriessen’s work this is intentional and a way of directing the listener toward the rhythmic and antiphonal features of the piece.
These harmonic aspects of the motive, while containing some interest in themselves, are subservient to the process of development the cell undertakes over time. When presented as part of the train “still-lives” (Train 1, Night Train, the coda of Building, and Spaceship), each of the five chords has a different meter which changes throughout the section.20 This is the process of primary musical interest:
20 In the knee plays, five maintains a constant meter within a repetition but changes meter between repetitions. This process as carried out in “Knee 4” is examined below.
The numbers in the table represent the number of eighth notes in each cell.
A few characteristics of the rhythmic process are immediately apparent. From
section B to K, with the exception of the fourth chord of F all rhythmic
processes are strictly prolonging. Each repetition chord is as long or longer
than the previous repetition of it. Another property of the process is that
each chord is held for either 3♪, 4♪, (3+3)♪, or (4+3)♪; that is to say, the
chords with 6 or 7♪ introduce no new figurations not heard in the 3 or 4♪
sections.21
21 Later presentations of five will be have lengths of 5♪ and 8♪ (Knee 3) and will
have 6♪ figures which are not literal repetitions of 3♪ (see "Knee 4,"
second cell, below).
The lengthening of chords (and thus cells) follows two distinct processes and
thus divide the section. The first cell, A (repeated three times) acts as an
introduction and a presentation of what becomes the standard form of the
motive—when five is heard at
the end of “Building,” A is the only form presented. The next five cells,
B-F, present the process of lengthening the motive from 3♪ to 4♪ beginning
with the fifth chord and progressing toward the first. Cell F alters the
process slightly by reducing the fourth chord to 3♪ while augmenting the
first to 4♪. Avoidance of the projection of regular meter within a cell seems
to be the overriding reason for this decision.
The next five cells, G-K, present a similar process, lengthening from 4 (or
3) eighth notes (via 6♪) to 7♪ beginning with the first chord and moving
roughly from front to back: 1, 2, 4, 3, 5. The cell which at B lasted 16♪ is
expanded by K to 33♪. The shift back to the quick transitions between chords
of L feels like a tightly stretched rubber band being suddenly released.
Without a change of tempo, the speed of the cell has been dramatically
increased and, with the return of the rhythmic profile of the introduction,
the process feels complete. By repeating L six times rather than three Glass
makes the coda more satisfying to the listener: while each chord is much
faster (3 or 4♪ rather than 6 or 7♪), by the fourth repeat (which does not
exist in any other cell) we are able to hear the cell not as a five-chord
motive but as two five-chord motives, a total of 36♪. Thus rather than
lessening the tension of increased cell length (B-K), L acts as a culmination
of this process. By focusing our aural “gaze” on two different levels of
activity, the pattern can be heard as both accelerating and broadening
simultaneously and without contradiction.
(The analysis of Einstein will continue in the a future blog post)
(Updated 2016-02: Found substitute YouTube videos; fixed links)
