Pattern Structure And Process In Piano Phase

Pattern Structure and Process in Steve Reich's Piano Phase PAUL EPSTEIN IN his 1968 manifesto, "Music as a Gradual Process," Steve Reich argues that once a composer constructs a pattern and sets it in motion by means of a rigorously defined process further intervention is unnecessary. Even when all the cards are on the table and everyone hears what is gradually happening in a musical process, there are still enough mysteries to satisfy all. These mysteries are the impersonal, unintended, psycho-acoustic by-products of the intended process.1 This statement represents a radical approach to musical form. Reich's

belief that music can be impersonal, unedited process, rather than an expressive, handcrafted metaphor for process, is a new idea as is the idea that the role of the composer is not as inventor of personal codes for us to decipher but as a discoverer of impersonal natural phenomena. In experiencing process music, the listener's task is also one of discovery—of the physical laws embodied in the process and of the psychological laws affecting the listener's interaction with the process. It is in this interaction that the coming together of impersonal and personal takes place that forms the key experience of process music. Few pieces by Reich or other composers actually present single unedited processes. At the very least, the open-endedness of many processes demands external intervention to terminate them. One of the most successful of the relatively pure process pieces, Reich's Piano Phase, composed in 1967, The preparation of tfaii e m y was supported by a grant from Temple University. It was first presented as part of a panel, "Minimalism: Cultural and Historical Perspectives," at the Annual Conference of the American Society of University Composers, Columbia, Ohio, April 5, 1984. At that time it was illustrated by examples generated on a Commodore 64 personal computer from a program written by Maurice Wright. The computer realization of Piano Phase was also used, together with tapes of live performances, in preparing the paper. This was done in order to ensure that discontinuities perceived in the phasing process were not the result of performance Inaccuracy. In fact everything described in the paper was verified in both live and digital versions. 1 Writings About Music (New York, 1974), p. 10.

494

Reich's Piano Phase

495

provides us with an excellent example of the surprises that can occur within a continuous process." The piece, written for two pianos but performable also by marimbas, is in three sections, with melodic patterns of 12, 8, and 4 sixteenth-note beats respectively. I will discuss only the first-and longestsection, looking at the inherent structural characteristics of the pattern, together with the phasing process by which the pattern is rotated against itself, as the source of the musical events of the piece. Ex.1.

The pattern (see Ex. 1) is actually a nesting of two repeating figures, one of three notes, the other of two. In addition to generating the melodic pattern, these figures function harmonically as the two contrasting sonorities of the piece. Piano Phase begins with the first player repeating the pattern, and that is all that player will do through the entire section. The second player enters in unison and after a while very slightly increases the tempo. Gradually the two separate until the distance between them is a full sixteenth note. Then the second player briefly matches tempo with the first before resuming the phasing process. The latter procedure continues through a full cycle of twelve phases, finally returning to close the section in unison. The second half of the cycle is a retrograde of the first, with the relationship between the two players reversed. (Note that in spite of different starting points, phases 1 and 11, 2 and 10, etc., are identical.) Two phases stand as formal markers: the unison that opens and closes the process, and phase 6, which appears only once, at the mid-point, and precedes the beginning of the retrograde. The unison situation is inherently basic, and its return constitutes a conclusion natural enough to the process to justify the claim of non-intervention. With the process defined as above, we have in effect a series of canons, at distances of from 0 to 11 sixteenth notes, alternating with transitions in which the two voices are out of phase with each other. Before examining the dynamics of the process, let me make several observations about the in-phase situations—phases 0 to 11 as shown in Example 2. First, the phases alternate between those in which the two sonorities, E-B-D and Ff-Cf remain intact—that is, each dyad is homogeneous—and

The Musical Quarterly

496 Ex.2.

v

6*

»•

L g .

r L • 9

" .

L

t-

0'

6F

r **>

o

'

* • mz

<•> 8

L 10

L"L"

«» . .

rr

those in which each dyad includes a note from either sonority. The evennumbered (homogeneous) phases are almost entirely consonant. With the exception of phase 6, they consist mostly of perfect consonances, with one pair each of minor thirds and minor sevenths. Phase 6 consists entirely of perfect consonances. The odd-numbered (heterogeneous) phases are significantly more dissonant, including three pairs of seconds, two of sixths, and one of perfect fourths. Looking at the consonant phases, it is predictable that at 0, 4, and 8 the F|'s and Cfs will be in unison; at 2, 6, and 10 they will be opposite each other, creating a static sonority on every other beat. The three-note sonority will only be in unison at phases 0 and 6. At 6 the interweaving of the unison E-B-D outline with the recurring Ff-C} dyad will separate the two sonorities in the clearest possible way. Another feature of phases 2, 6, and 10 follows from the structure of the pattern. Since the three-note sonority is stated exactly twice during the pattern, and since in phases 2, 6, and 10 an F | will always be opposite a C | (thus neutralizing the duple structure), these phases will consist of

Reich's Piano Phase

497

a repeated six-beat composite pattern. This will always happen with a canon at a distance of half the length of the pattern, but it is unusual at other distances. Phase 5 presents a startlingly unexpected situation, repeating dyads. But this is also a direct consequence of the patterns structure. It results from the fact that in the original pattern two-note successions starting at note 1 are reversed starting at note 5 (see Ex. 3). This in turn follows from the displacement, in the second half of the pattern, of one sonority in relation to the other. Phase 5 is perhaps less surprising in itself than in the way it emerges during the phasing process, and we will return to it in that context Ex.3.

All the musical cards are now on the table. We have a rigorously defined system, a process that is continuous and free of external intervention for expressive, formal, or other aesthetic reasons. Where, then, are the mysteries? They are in fact numerous and stem in part from the fact that while the process is continuous, our perception of it is not. The listener is presented with a rich array of possibilities out of which he/she may construct an experience of the piece. The events that constitute the real musical material of Piano Phase may be divided into those that occur within a stable situation and those that arise in the course of phrasing. The first include composite subpatterns and reconfigurations; alternative accentual or metric interpretations of the material; and the submerging and reappearance of the original pattern. They may seem to impose themselves on the listener; or the listener may actively cultivate them, deliberately focusing on a particular reading or even going back and forth among several alternatives within a single situation. Events during phasing concern when and how the change from one phase to the next is perceived, as well as with the nature of the change. The first option is presented within the pattern itself. The pattern structure described earlier could have been applied to any three-note and two-note groups. One aspect, however, is specific to the actual pitches selected. Because of the gap between Ff and B, the pattern tends to divide

The Musical Quarterly

498

registraUy into two voices, as shown in Example 4. The nature and prominence of the two-voice texture varies in the course of the piece. In some phases it is strikingly polyphonic (see Ex. 5); in others, as in phase 2, shown in Example 6, the lower voice supports the upper homophonically; and in several—notably phases 1, 5, and 6-the voices do not easily separate, or other configurations supersede the two-voice structure (see Ex. 7). Ex.4.

y

pITT

Ex.5.

Ex.6.

Ex.7.

The phasing process begins with a movement away from unison. Although continuous, it is heard in several distinct stages. At first the impression is of increasing resonance, a change in acoustic quality only. At the next stage one begins to hear the voices separate: echo replaces resonance. At a certain point the irrational division of the beat caused by the echo presents.a dizzying rhythmic complexity. When the voices

Reich's Piano Phase

499

are nearly 180°, or one half beat, out of phase, a doubling of the tempo is perceived; one has a momentary sense of stability, of a simplification of the irrational rhythmic relationship heard previously. This stage is very brief and is one of those events that seem to occur suddenly. The outof-phase quality quickly returns and lasts until the new phase locks in. Phase 1 is heard in effect as a continuation of the echo. The fast tempo and the identical range and timbre of the two instruments prevent the canon from being heard as such. Yet several factors combine to make the pattern itself audible—notably, the frequent major seconds, as a result of which the contour of the pattern is largely intact, if somewhat blurred; and the absence of any significant composite structure or repeating subgroup such as will help to mask the pattern in later phases. In this regard, notice that while each dyad occurs twice, no succession of dyads is repeated. - A very different situation occurs at phase 2. Here the dissonant rubbing of the pattern against itself is replaced by a consonant texture in which, as we have seen, the static Ff-Cf alternates with a dyad from the E-B-D sonority. Furthermore, the measure consists of a six-beat figure that is repeated. And, most tellingly, the two-voice division of the original pattern is fully realized. This results in a placid, consonant sound with a stepwise rocking figure in the upper voice accompanied by a syncopated lower voice. The simplicity of this texture, and especially the internal repetition, masks not only the canonic relationship but the pattern itself. Oddly enough, the pattern is further obscured by the existence of two additional canonic voices. As shown in Example 8, these are formed by alternate notes of the actual voices starting at the fifth and seventh notes of the measure. The pattern is thus submerged in and obliterated by its own reflections. Ex.8.

$

The real significance of the extreme contrast between phases 1 and 2 lies in the transition between them. Again it is the abruptness with which certain changes seem to occur that is startling. In this case there are three stages: we hear phase 1, with the two voices gradually moving apart; near the midpoint the notes begin to swirl around chaotically, and it becomes increasingly difficult to follow the pattern; but a moment later, the new

500

The Musical Quarterly

configuration begins to coalesce, usually well before the instruments actually synchronize again. There may be considerable variance in the listening experience from one individual to another and from one hearing to another, in how the exact moment of changes is perceived. It is even possible, with practice, to choose to retain the old pattern a bit longer or jump to the new a bit sooner. And during the intermediate stage it is briefly possible to alternate between hearing and not hearing the pattern. But as soon as one hears the high-low two-voice division of phase 2, whenever that may be and whether inadvertently or deliberately, the old pattern is irretrievable. The transitions between phases vary in both content and character depending on the nature of the preceding phase, the following phase, and the composite double-time melody formed at the midpoint of phasing. I will only examine one other transition, from consonant to dissonant: phase 4 to phase 5. Phase 4 contains a highly polyphonic two-voice texture. But because the F}'s and Cf's are in unison, it is also possible to discern the original pattern and perhaps even the canon—or at least to focus on either of the canonic voices. In phase 5, as in phase 1, one can hear a rough paraphrase of the pattern; but any more complex structures are overshadowed by the insistent repetition of the dyads. The progression from consonant to dissonant is then, strangely enough, one of textural simplification. The key to the phasing from 4 to 5 is the separation of the unison Ft's and C$'s into repeated notes (see Ex. 9) and their recombination to form the repeating dyads. The repeated notes permit us to follow the phasing process much more easily than in some of the other transitions. A good way is to focus on either the high or low voice in phase 4 and follow its progress. Ex.9.

$

^ £

An additional area of potential material in this music is opened up by the nature of meter. In much of the music of pattern repetition, meter is largely a quantitative factor; one hears that a figure repeats after so many beats, but there is no functional downbeat—or several accents may compete for primacy. (I have heard musicians in Reich's ensemble speak of the difficulty, in performing his music, of "finding the 1.") This non-hierarchic

Reich's Piano Phase

501

meter helps to propel the music forward by eliminating the heavy articulation of an unambiguous metric accent. In Piano Phase the pattern itself is fairly unambiguous, but once, phasing begins the metric situation becomes more complex. Then it is quite likely that either momentary inattention or the emergence of a new subpattern will result in a metric reorientation. Even a reversal of strong and weak beats is possible—especially in odd-numbered phases, where harmonic alternation is absent. Example 10 shows a metric reorientation of phase 3. Ex. 10.

One may adopt one of several strategies as listener, from passive to active. The former seems advisable when this music is first heard. Later new or more remote possibilities may be sought through metric reorientation and other procedures, including a selective focusing on certain notes and ignoring some others. Example 11 shows a pattern derived from phase 2. The dotted line represents the original barline. Ex. 11.

In "Music as a Gradual Process," Reich speaks several times of the "impersonal" quality of the events in his music. In fact the experience of listening to Piano Phase greatly resembles observing a gradual process in nature. Especially comparable is the appearance of discontinuities. For example, several such phenomena mark the final stages of a total solar eclipse. Here is a description from 1836 by the British astronomer Francis Baily of what came to be named "Baily's Beads": When the cusps of the sun were about 40° asunder, a row of lucid points, like a string of bright beads, irregular in size and distance from each other, suddenly formed round that part of the circumference of the moon that was about to enter on the sun's disc.

502

The Musical Quarterly

Its formation, indeed, was so rapid that it presented the appearance of having been caused by the ignition of a fine train of gunpowder.3

In the music, discontinuity is of several kinds. During phasing the ear will identify certain discrete landmark situations—the splitting of a unison, the doubling of tempo at the midpoint. Even though it is apparent that these have been arrived at gradually, the ear identifies them within only a narrow margin of error, and this results in a feeling of abrupt change. Sudden perceptual shifts occur when phasing has progressed to the point where one configuration is no longer viable and is replaced by another. In such cases the discontinuity is purely perceptual, the actual change being merely one of degree. As we have seen, perceptual shifts—voluntary or involuntary-also take place in the total absence of change in the music. The likelihood of such shifts, which are akin to figure-ground reversal, increases in proportion to the degree of ambiguity of some aspect of the music. I have suggested that the unison phase of Reich's work has special significance, and thus it is appropriate to end with a brief look at its return at the conclusion of the phasing process. A second event associated with a solar eclipse provides a striking parallel to the return to unison. The moment of total eclipse is notable for the sudden appearance of the solar corona, which at all other times is masked by the sun's brightness. Similarly, the reappearance of unison in Piano Phase is such an event because, as the pattern emerges and finally locks into phase, we are reminded that, however obscured, it has been there all along.

1

Quoted in Agnes Cleike, A Popular Hixtory of Attronomy during the Nineteenth Century (London, 1902), p. 61.