Article

Concepts of Rhythm in the Performance Practice of Twentieth Century Music

Mieko Kanno

Lecture given at Performance Practice session of SECOND BIENNIAL INTERNATIONAL CONFERENCE ON TWENTIETH-CENTURY MUSIC Goldsmiths' College 28 & 29 June 2001 (PGW)

Rhythmic accuracy is largely considered to be one of the most critical principles in the performance of contemporary music. 20th century composition achieved a high degree of technical sophistication which yields a vast number of rhythmic and metrical ideas to be expressed and differentiated from one another.

Such advancement subsequently changed conventions of performance and made almost redundant some of the more traditional interpretative practices which we still observe in pre-twentieth century music. Consequently, 20th century music altered our attitude towards performance practice to such an extent that we deploy two distinct categories of interpretation for old and new repertoires: it is typically thought that 20th century music requires more accuracy regarding notated rhythmic values, whilst the practice of the older repertoire continues to employ agogic rhythm as a critical tool in performance. In the older repertoire this principle is also applied to matters of metrical structure, such as tempo and phrasing, and these remain firmly under the interpreter's authority.

In the climate of increasing sophistication, the neutral approach towards rhythm on the part of the performer may have been a necessity in the course of satisfying the required articulation, which can be extremely complex or so removed from conventions that the performer is left wondering how it can ever be realised. However, it remains to be asked whether such neutrality really benefits the expressive purposes of the rhythmic and metrical structures of 20th century musical works.

I want to discuss this question: whether a neutral approach towards rhythm befits the music or not, first in the context of the viola piece, Coelocanth (1955), by Giacinto Scelsi. Here the notation has no bar-lines, no time signatures, no signs to tell us immediately what rhythmic and metrical structures govern this music. The only information we have regarding these structures is that the rhythms are organised in groups of crotchet length, and that the crotchet length - at the metronome mark 60 - constitutes the beat.

Coelocanth (1955) for solo viola by Giacinto Scelsi,
the beginning of the first movement

Let us start with examining this given information. Metronome marks function as a real time reference in music. This manner of reference, the pinning down of musical time to real time, can be observed in space-time notation too. Either a clock or a two dimensional line gives the standard against which music can be mapped out. Then, we see that a mapped-out performance of a piece under metronomic guidance identifies the temporal structures of that piece with those of the mechanical clock, and that a mapped-out performance of a piece in space-time notation identifies the temporal structures of that piece with the physical constraints of a two dimensional line. These facts imply that the teleological sense of the past, present, and future of the referred parameter (the mechanical clock or the two dimensional line) is in complete agreement with the expressive purposes of the music. Understanding this type of time relationship is vital for the articulation of some music: much commercial music benefits from such performance practice, in which the tune surfs on top of the underlying beats, taking advantage of the regularity of the beats as a motor to create momentum in music. But, would such an automatic transcription of given rhythms as sounds along the metronomic beat express most musics as appropriately? Would a mobile-phone style transcription of this Scelsi passage be considered successful? If not, what other kinds of rhythmic approach does this music embrace?

In performance, a mobile-phone style transcription of this passage lacks any sense of direction, even after adding dynamics. The music appears to encourage some form of characterisation through rhythm or phrasing; it appears directionless and unnecessarily long when it is played metronomically and according to the outlines given in the score.

A closer look at Scelsi's distribution of notes provides an insight: the sustained melodic contours appear to suggest continuity whilst registral shifts of the pitch parameter give an impression of gestural interjections. There are pitches, particularly among the long held notes, that appear to have a nodal function in the harmony. Also, the number of long held notes increases towards the end of the movement. From these points made by observing the score, we may arrive at an understanding that these held notes articulate the basic structural pillars, around which all the other quick gestures are interspersed. The movement can be seen to evolve around the static plateau of these held notes.

But how can we make sense of such a stasis in performance? There are at least two groups of possibilities for interpretation. The first group is that the performer can create a flow of music by means of emphasising gestural patterns and phrases, including rubatos if necessary, thus interpreting the music like a Baroque fantasia. Long held notes can be understood to represent phrasing punctuations. This type of interpretation produces a momentum in performance, makes the passage flow better, and thereby alleviates its directionlessness.

The second group of possibilities requires a more fundamental change in the position from which one can approach music. The stasis in this passage is caused by its frequent retreat into long held notes, which deprives the quick gestural passages of any continuous development. Such static organisation of events can be attributed to a particular temporal mode of expression, in which the gestural materials are self-referential and never meant to develop. From this point of view, we arrive at a very different understanding about the way in which rhythm can be articulated in musical time. This temporal expression is based on a distinct concept of time: time as a space, as a non-developmental plateau in which materials articulate themselves against the flow of time. We know from experience that music can create its own time, that the temporality we experience in music isn't necessarily identical to real time. Scelsi's long notes can be seen to exist in a frozen time, where the progressive movement of time is replaced by the accumulative process of time. From such an understanding of temporality, the concept of rhythm emerges as having quite a different function in this music. Here rhythm can be seen as the consequence of a non-linear, spatial movement unfolding in time. It follows from this that rhythm, in this sense, doesn't express itself in the metrical grid, but expresses itself through the metrical grid.

The question of which one of these two approaches is more appropriate to the performance of this music, is another matter. There remains another approach, the neutral mobile-phone style, which some musicians may prefer. But this example suffices to indicate that interpretative sensibility towards musical rhythm can have distinct roles in articulating the rhythmic and metrical structures of a musical work in performance. The point is that even such a simple concept like phrasing and direction, derived from rhythmic patterns on the page, can be articulated in multiple ways. The aforementioned neutral approach to music in performance, in which one maps out sound according to the beats, emphasises accuracy of rhythm from the point of view of metrical structures in music. Despite the merit of universal correctness in the result, this approach is often ever so close to trivialising the multi-dimensional nature of the expressive power of rhythm in music.

I have mentioned that Scelsi's music can be understood to explore the spatial dimension of time. As many musical works which we enjoy today demonstrate, composition in Western art music has a strong inclination towards a spatial contemplation in which time is given a spatial element. This may well be to do with its notational practice, which is a spatial representation of what we consider to be music. Such conceptual independence of musical time from the flow of real time often plays a vital role more prominently in 20th century composition. Jonathan Kramer considers this inclination on the part of composers to be indicative of non-linear structuring of music; he observes the impact of the spatial, non-linear concept of time in many 20th century works in the perception of which a musical present acquires a distinct meaning of its own. We could cite many 20th century works as examples of this, including Stravinsky's as well as a large proportion of works composed after the mid-century. This non-linear manner of expression of time often includes atemporality as an essential part of its expressive mechanism. Temporal non-linearity can be understood as a non-developmental manner of discourse, and its primary expression unfolds in space rather than in time.

In contrast to the increased possibilities for varied rhythmic and metrical structures in 20th century composition, it is ironic that such compositional variety and specificity severed the very potential of expressive variety in performance practice. Indeed the variety of rhythmic articulation in performance decreased in inverse proportion to the increase of the same articulation in composition. The suggestion that the expressive richness has moved from performance to composition doesn't really hold water, because the performance more frequently results in a cancellation of the richness the composition attempted to articulate. There is nothing to prevent spatial temporality in composition from broadening the expressive range of rhythmic and metrical structures in performance too. But we cannot dismiss, on the other hand, the allegation that the neutral approach in current performance practice is rooted in the spatial compositional premise, in that the more an atemporal approach dominates composition, the more confusion over concepts of rhythm increases in performance.

Performers' confusion over concepts of rhythm can be attributed not only to its diversity in composition, but also to the very nature of rhythm. So far I have been using the two terms, rhythm and meter, without any qualification. However, the distinction between rhythm and meter may be too critical to be taken for granted in the context of discussing performance articulation of the linear and non-linear temporalities. Our simplistic yet practical understanding of meter and rhythm is that rhythm is a rich and fully sensuous embodiment of music's temporal progress, and meter is abstract, mechanical, devoid of any intrinsic expression, and is thus rhythm's shadowy, schematic counterpart. Christopher Hasty encourages a better understanding of the relationship between them in his book Meter as Rhythm: he argues that rhythm is an essentially more musical phenomenon than meter, and proposes that meter can be considered entirely as rhythm for a musical understanding of a work.

There are many similarities and differences between rhythm and meter. The two have so much in common that neither of them can exist in music entirely without the other. One critical difference is the following: meter is understood as a dividing agent, presupposing the existence of a whole, be it a section or work. On the contrary, rhythm is understood as an adding agent, orientating itself as the only concrete material and leaving the possibility of a whole as a probability. The difference between rhythm and meter is in the way in which they relate to the whole, either as an embryonic cell from which the whole proliferates, or as a facet of the pre-existing whole.

This distinction invites us to observe that there are shared elements between the opposition of rhythm and meter on the one hand, and the opposition of linear and spatialised approaches towards music on the other: the linear approach towards music has only the present moment as its unit, to which the past and future moments are chained to create a duration of time. On the contrary, a spatialised approach towards music takes the framework of duration as its unit. Here time is usually organised as a finite space with a beginning and end, thus it is rendered as a segmentable parameter. From this, one can deduce that rhythm is the most basic parameter in which linearity of music can be expressed, and meter is the most basic manifestation of the organised whole which can be expressed in space.

However, as in the case of meter and rhythm being interdependent, there are no musical compositions that are either totally linear or totally spatial. Even the most straight-forward sounding piece of music has some spatialised elements of structure. Even a simple refrain is already a spatialised structure; nothing is repeated identically the same in real time, but repetition forces this to happen, as closely as possible, by directly linking to the juxtaposition of identical materials in space. Repetition is one of the most fundamental techniques of spatialisation. Indeed, one can say that music is music because it has more than a single dimension of time. At the far end of this spatial evolution of music lies the graphic score: despite the fact that it is a spatial presentation in itself, its realisation unfolds in time. Here music wouldn't be music, if it didn't express itself in the linear dimension of time. Interaction between the two is critical in the expression of musical time.

It follows from the foregoing argument, that rhythm and meter do interact, in that the additive mode of rhythm intermingles with the divisive mode of metre to create musical time. Assessing the ways in which musical composition integrates these two modes of temporal structuring can lead to a better understanding of how to articulate rhythm in performance. How to articulate rhythm, rather than meter, precisely because of rhythm's proximity to musical articulation in performance, as pointed out by Hasty. Many musicians as well as musicologists have observed the way in which Stravinsky combines the expressive potential of the interaction between rhythm and meter to effect a more varied sense of temporality in the performance of his music. The earlier example of Scelsi shows preliminary ways to deal with this in performance.

Finally, I want to consider briefly the topic of how our considerations about rhythm relate to the current obsession about accuracy in performance. My discussion of rhythm and meter leads to a proposal that accuracy lies within the interaction between the outlined scheme in composition and the actual circumstances of its articulation in performance. These two seldom overlap. But this isn't in any sense a disadvantage, because the friction between the two constitutes what music is. The tension between the two is reflected in the tension between meter and rhythm, between linear and spatialised temporalities. The next question is how effective this interaction can be, and how musical the attained accuracy is in the context. For the performer, this is a question of how to translate the given information, which is often a set of metrical and spatialised ideas and details, into the linear dimension of real time. This question can be raised equally in very complex works as well as simple ones (such as Scelsi's), because the tension between heterogeneous elements is omnipresent in music. The question of computability of these elements as information is at the heart of this practice of accuracy in performance: it is a question that concerns both composers and performers, and also an area no less critical for musicologists.