Glass the Traditionalist: Subjectivity and Teleology in Einstein on the Beach

by Kevin Hartnett

In the late 1960s, Philip Glass and several of his contemporaries—among them Steve Reich, Terry Riley, and La Monte Young—rebelled against the dogma of the academic musical establishment and created with modernist zeal a music that would soon be described as “minimalist.”  Due to its widespread influence in the years since, it is perhaps now difficult to imagine just how radical and innovative this music was for its time.  After decades of esoteric chromatic serialism, the unabashed consonances, steady pulses, and relentless repetition of this new “minimal” music marked a bold departure from the familiar sound-world of the avant-garde.  Additionally, the early music of Glass and his cohorts functioned completely outside of the tradition of conventional dialectical Western music.  Instead, its emphasis was on process and objectivity.  In his 1980 review of the American minimalist movement, Belgian musicologist Wim Mertens characterizes the difference as such:

Traditional dialectical music is representational: the musical form relates to an expressive
content and is a means of creating a growing tension; this is what is usually called the
“musical argument.” But repetitive music is not built around such an “argument”; the work is
non-representational and is no longer a medium for the expression of subjective feelings.

Mertens is not alone in describing minimalist music in these terms.  Steve Reich’s widely-read 1968 essay “Music As a Gradual Process” outlines the composer’s preference for audible, impersonal processes.  Philip Glass later echoed Reich’s sentiment and suggested that “the listener will therefore need a different approach to listening, without the traditional concepts of recollection and anticipation.  Music must be listened to as a pure sound-event, an act without any dramatic structure.”

This clear ideology expressed with both words and music successfully fueled the development of Glass’s music for nearly a decade.  To achieve this desired objectivity, Glass frequently employed clear, rigid mathematical processes that would later be inextricably linked to his musical vocabulary.  However, by the mid-1970s, Glass began to find continued adherence to this strict brand of objective radicalism to be unsustainable.  “To break the rules so rigorously is actually to follow them unfailingly,” he said in a 2009 interview with critic Tim Page.  “I just turned black into white.  That’s all I did.”   With the completion of his deliberately didactic Music in Twelve Parts in 1974, Glass felt he had reached the limits of the austere minimalist aesthetic found in his earlier works.  It was around this time that he had begun an ambitious collaboration with the experimental stage director Robert Wilson under the working title Einstein on the Beach on Wall Street.  The music written for the resulting four-and-a-half hour-long opera—its title later shortened to Einstein on the Beach—would come to mark a notable point of maturation in the evolution of Glass’s musical language.

At a cursory glance, Philip Glass’s music for Einstein on the Beach may appear to be strikingly similar to his preceding music; it remains rife with the repetitions and pulses that many had come to expect from the composer, and it was written for the same unique ensemble, the Philip Glass Ensemble, that had been presenting Glass’s music in downtown Manhattan for nearly a decade.  Upon closer examination, however, it becomes clear that in this piece Glass broadened his technique and freed himself from the strict mathematical processes he had employed in the past.  Composer and musicologist Kyle Gann points out that Glass’s treatment of musical material in Einstein is “far from arithmetically simple” and that it is sometimes even “devoid of predictable algorithmic thinking” altogether.  “Looking back in retrospect,” he goes on to write, “Einstein seems a far more intuitively written work than we thought at the time.”  Glass’s move towards subjectivity—something he had once aimed to eradicate from his work—marks a striking departure from his previous philosophy and the prevailing objectivist paradigm of the time, and the result is a music that is capable of expressing exactly what Wim Mertens claims his music lacks entirely: a “musical argument.”  In this paper I will show that, contrary to Mertens’s claim, Philip Glass’s “repetitive music” in Einstein on the Beach is indeed built around a musical argument.  Its underlying teleological flow is achieved not through the traditional manipulations of the relationships between harmony, tonal centers, and conventional forms but by analogous manipulations of meter, pulse, and recombinant forms.  Glass’s control of these musical mechanisms is perhaps most clear in the “Night Train” scene of Act II, the scene on which my analysis will focus.

To understand Philip Glass’s treatment of meter, pulse, and form, it is first necessary to examine various aspects of pulse and meter and ways in which they can be perceived.  In order for the listener to perceive a pulse, there must be a series of attacks or articulations that occur at regular intervals.  The pulse is then established once the listener is able to recognize the regularity of these events.  Glass establishes the pulse at the beginning of  “Night Train” with a repeating organ figure (figure 1).  A steady flow of eighth notes continues in the organ for the entire 20-minute duration of the scene, leaving little room for pulse-related ambiguity.

The establishment of clear metrical organization operates in a similar fashion; it too is dependent on a series of recurring musical events— such as patterns in pitch, contour, dynamics, or articulation—that organizes the pulses into regular groupings.  The only difference in this case is that meter, by definition, operates on a larger scale than pulse.  Theorist Maury Yeston calls these two levels of rhythmic organization “rhythmic strata.”  He claims that the perception of meter is “an outgrowth of the interaction of two levels—two differently-related strata, the faster of which provides the elements and the slower of which groups them.”  In his response to Yeston’s work, Harald Krebs takes this concept a step further.  He points out that all levels of motion will operate either as a “pulse level” (the fastest level) or as one or more “interpretive levels” (slower levels which suggest metrical groupings).  He uses the term “cardinality” to describe the number of pulses between each attack of an interpretive level.  For example, the recurring gesture of four ascending eighth notes that is repeated incessantly by the organ (figure 2) could suggest an interpretive level with attacks every four eighth notes.  Thus the cardinality of this interpretive level, with four pulses between its own attacks, is 4.  According to Krebs’s nomenclature, an interpretive level with a cardinality of 4 would be labeled “4-level”; I will continue to use this terminology throughout the remainder of my analysis.

The implementation of two or more strata presents the opportunity for a rich variety of interactions between them.  The most simple of these, as Krebs points out, are relationships that can be described as metrically consonant.  These consist of passages in which the attacks of each interpretive level correspond with attacks in every faster level.  The measure in which the bass clarinet and vocal soloists enter, for example, can be thought of as metrically consonant (figure 3).  The repeated figures sung by the soprano and tenor have a duration of eight pulses therefore creating two 8-level gestures.  As we have established, the organ part (specifically the left-hand part, now complemented by the bass clarinet) suggests a 4-level interpretation.  When combined into three strata, it becomes apparent that all large-scale attacks of the 8-level components coincide with attacks of the more-frequent 4-level component as well as with the pulse level.  This degree of agreeability, or consonance, between strata is responsible for the relatively low sense of tension conveyed in this measure.

The relationships become more complex in the following measure.  The left hand of the organ and bass clarinet remain locked in ostinato, but the soprano, tenor, and right-hand organ gestures have been shortened to six pulses (figure 4).  The attacks of the voices’ stratum, now transformed into a 6-level, do not coincide with every attack of the organ’s 4-level ostinato; therefore, we can conclude that this measure is metrically dissonant.  Additionally, the voice figure must repeat four times and the organ and bass clarinet ostinato three times before the figures return to their original alignment.  The length of this complete cycle, or phase, can be useful in describing the degree of dissonance between strata.  In general, the longer the phase length, the more dissonant its constituent strata.  At 24 pulses, the phase length in the measure shown in figure 4 is much longer than the eight pulses of that shown in figure 3.

In addition to the vertical, or direct, metrical dissonance described above, Krebs also describes a dissonance that occurs linearly when dissonant interpretive levels are juxtaposed rather than superimposed.  This “indirect metrical dissonance” can be found when the aforementioned excerpts are examined in context (figure 5).  To illustrate, the 8-level cardinality of the voice parts in measure 3 is followed directly by the 6-level figure in measure 4.  This jarring switch of metrical emphasis and the resulting direct dissonances with the 4-level ostinato work simultaneously to heighten the sense of tension.

Moments of indirect dissonance play an important role in shaping the large-scale form of “Night Train.”  While Glass frequently employs moments of indirect dissonances in the vocal parts, he uses them much more sparingly in the left hand of the organ.  With the exception of the highly unpredictable introduction section and the gesturally distinct coda, the organ departs from its 4-level ostinato only five times.  In these moments, it shifts suddenly to a similar figure with a 3-level cardinality and stays there briefly before returning to the familiar 4-level pattern (figure 6).  These indirect dissonances in the organ are particularly striking for three reasons.  The first reason, as was mentioned above, is that they occur only rarely.  Secondly, they spring suddenly from what was a predictable and reliable ostinato.  While the organ is generally used as a reference point against which other consonances and dissonances are measured, its momentary lapse into a conflicting cardinality marks an intentional and dramatic change in tone.  Lastly, and perhaps most importantly, they usher in brief but drastic changes in texture that divide the piece into five clear sections.

These five sections, in turn, shape the scene into a reiterative form that Kyle Gann calls “stanzaic form.”  The stanzas, which are similar in motivic content, are separated by solo organ interludes.  Gann also identifies particular motives as having specific introductory and concluding functions that he calls incipits (organ, measure 18) and envois (voices and organ, measure 16), respectively.  These motives reliably book-end each of the stanzas, and they can help the listener anticipate when one stanza will end and when the next will begin.  In this way, the incipit and envois motives have a cadential function.  Each envois occurs before the voices drop out of the texture; furthermore, the shortened cardinality of the left-hand organ part—3-level as opposed to 4-level—is analogous to the accelerated harmonic rhythm that is typically found near tonal cadences.  The resulting increase in tension is carried over into the interlude before finally resolving into the 4-level incipit motive (which also marks the return to the ostinato).

An important feature of this stanzaic form is the fact that its reiterative nature allows for the repeated recomposition of material.  Rather than follow the exact same path through each stanza to reach a consistent goal (the envois motive), Glass uses recombinant musical elements to alter the course to the goal, lengthening the route with each new stanza.  According to Gann, Glass’s use of recomposition in “Night Train” makes each stanza “parallel in their function and similar (though varied) in their progress through the same harmonies and motifs.”  The relative length of each stanza can be seen in figure 7, where each column of numbers denotes a single measure.  For example, the first stanza contains five columns which corresponds to five measures (including their embedded repetitions).  When comparing these five stanzas to each other, it becomes clear that Glass is intentionally creating a sense of delayed gratification by incrementally—though not gradually or systematically—increasing the amount of music it takes to reach the end of the stanza.

Glass also mimics traditional methods of desire creation by controlling the degree to which the voice parts stray from the primary metrical consonance.  As mentioned earlier, the phase length (in pulses) can be a useful way of quantifying a degree of direct metrical dissonance.  In the first stanza, the highest number of pulses per phase is 24.  In the second stanza, however, Glass introduces an even more dissonant pair of measures, each with a phase length of 40 pulses (figure 8).  There are no phases of 40 pulses in the third stanza, but they return in an even more prominent role in the fourth stanza.  Glass’s use of distantly related metrical dissonances is directly analogous to the traditional movement to distantly-related key areas in tonal music; again, his controlled but non-systematic approach to controlling these instances of dissonance suggests a deeply intuitive compositional process.

In addition to manipulating the frequency and degree of direct metrical dissonances to shape a large-scale musical structure, Glass also carefully employs moments of indirect metrical dissonance to strengthen his musical argument.  The music rarely stays in a position of consonance or dissonance for more than three measures; the frequent indirect dissonances are a result of the constant swing between consonant and dissonant measures.  These swings are most pronounced in compound measures (figure 9).  Compound measures, unlike their simpler counterparts, feature two repeating musical cells within the measure rather than one repeating cell.  Each of these cells oftentimes has a different cardinality, resulting in an indirect dissonance within the measure.  Compound measures maximize the impact of indirect dissonance by switching quickly between their opposing constituent musical cells, highlighting the clash between them.  The effect is that of severe metrical disorientation; there is little time for the listener to catch his metrical bearings before the cardinality is abruptly switched.  Just as the increasing frequency of direct dissonances creates teleological flow, so too does the increasing frequency of compound measures. They first appear in the third stanza, and they become more and more prevalent towards the end of the scene.

With all of these tools at his disposal, Glass is able to steadily increase the amount of metrical tension—and by extension, dramatic tension—as “Night Train” progresses.  In the interlude preceding the fifth and final stanza, he spends an unprecedented amount of time playing with material related to the envois and incipit motives which heightens the level of anticipation and desire for metric resolution (figure 10).  The cardinalities of the organ figure change more rapidly than ever before, and the swirling gestures are denied metrical resolution for a full 25 seconds.  When it finally resolves, the incipit is been truncated to half its normal length.  It is at this point that the full ensemble, which now includes flutes and chorus, suddenly rips through texture to begin the climactic final stanza.  It is in this stanza that we see the greatest saturation of direct dissonances and compound measures.  The last two measures of the stanza in particular make up a point of significant tension (figure 11).  While it is not the first time that the envois motive appears within a compound measure, it is by far the lengthiest occurrence (nearly 40 seconds) of the motive.  Just as he did with the final interlude, Glass maximizes the desire for resolution of the envois by delaying it for an unprecedented amount of time, therefore solidifying the stanza’s status as the crux of his musical argument.

The dramatic distribution of these mechanisms that control tension—dissonances, delayed resolutions, changes in texture—makes it difficult to justify Mertens’s claim that repetitive music, including that of Philip Glass, is not built around a musical argument.  Though “Night Train” and other scenes from Einstein on the Beach are void of the conventional mechanisms that control dramatic tension, they are rich with alternative mechanisms that achieve the same effect.  In this way, the work remains innovative yet deeply rooted in the tradition of Western dialectical music.  Upon hearing Einstein for the first time in its entirety in 1981, composer Tom Johnson wrote that the work “may be minimalism in a kind of sociohistoric sense, but it has little to do with the purer minimalism of other composers, or with the spirit of reductivism so widely practiced in the visual arts.”  So if he is not a minimalist, how do we describe Philip Glass?  Is he a Classicist?  A Romanticist?  “I really can’t figure out what to call him,” Johnson concludes, “other than a good composer.”



Gann, Kyle. “Intuition and Algorithm in Einstein on the Beach.” NewMusicBox, March 6, 2013.

Glass, Philip. Einstein on the Beach. Dunvagan Music Publishers, Inc., 1976.

Glass, Philip. Interview with Tim Page. The Forum@MC. University of California Television. February 18, 2009.

Krebs, Harald. “Some Extensions of the Concepts of Metrical Consonance and Dissonance.” Journal of Music Theory Vol. 31, No. 1 (1987): 99-120.

Johnson, Tom. “Maximalism on the Beach: Philip Glass” The Village Voice, February 25 – March 3, 1981.

Mertens, Wim. American Minimal Music. London: Kahn & Averill, 1983.

Yeston, Maury. “The Stratification of Musical Rhythm” Journal of Music Theory Vol. 21, No. 2 (1977): 355-373.