Read an Excerpt
A Theory of Musical Semiotics
By Eero Tarasti
Indiana University PressCopyright © 1994 Eero Tarasti
All rights reserved.
IN SEARCH OF A THEORY
1.1 A Brief Critical History of Musical Semiotics
An implicit starting point for the study of all cultural phenomena is the fact that we live quite literally surrounded by various signs and significations, at the intersection of messages coming from everywhere. This universe of signs includes musical signs and messages. Not only are all stylistic periods of musical history simultaneously present in our time, but also our musical consciousness has expanded in an anthropological dimension, as Westerners are increasingly exposed to the different values and realities of non-Western cultures. Against this background it seems natural that musicologists, too, should join the search for theories and models that can explain, more effectively than before, the total heterogeneity of musical reality. Partly because of these circumstances, an entirely new branch of research — musical semiotics — has emerged; and interest in this type of research has also increased considerably among scholars representing "traditional" musicology. The dispersion of our musical reality justifies the question whether there exist any common features in all human uses and practices of music, i.e., musical universals, categories of the human mind that form a basis for any and all musical activity.
The problem can be put another way: Is it possible to construct a theory, a musicological metalanguage, to deal with and describe these universals? More particularly, can musical semiotics serve as this kind of integrating method? The first problem encountered, considered crucial by Charles Seeger, concerns the gap between two kinds of knowledge about music (1960: 225). According to him there is a conflict between the inner knowledge of music (felt immediately by producers and receivers of music) and the knowledge conveyed verbally, "outside" the musical process. Seeger viewed the relation between the inner universals of music and the speech universals of its observation from the "outside" as the basic problem of musicology. How can musicology, which is verbal activity, bring knowledge of the inner logic of music?
This question can be answered in three ways. First, one may assume that the two sets of universals overlap, that there is no basic difference between them, and that language can express the most essential aspects of music. This viewpoint is an oversimplification, for if verbal and tonal languages were parallel phenomena, we could not explain why musical expression has developed so far and formed a specific discourse in almost all cultures. The second viewpoint is supported by many musicians, who claim that musical and verbal knowledge are mutually exclusive. Someone may have a wide verbal knowledge of music without any idea of how music sounds or is made to sound. To speak about music is useless, suffice to play it. This attitude ignores facts of music history, since musical tradition is essentially transmitted with the support of verbal expressions, and no sign system in any culture functions separately from other modeling systems. Therefore, the third and most reasonable alternative would be to take a mediating position, from which we can say something true about music while accepting that some aspects of music cannot be depicted verbally, though some aestheticians paradoxically have written whole books about music and t he ineffable (e.g., Jankélévitch 1961). Consequently, musicology must consider its place among different ways of speaking about music and the relation of its metalanguage to musical phenomena, and such consideration must include semiotical observations concerning the nature of the musical sign system.
Sign systems can be classified according to their physical channels or modes: whether their expressions belong to the visual sense (gestures, mime, writing, plastic arts, etc.) , the auditory sense (spoken language, music, etc.), or the tactile sense (t he language of the blind). We should note that musical reality is not limited to aural expressions. As Ingmar Bengtsson has suggested (1977: 17), the concept of tone can refer to notation (tone n), or to a physically measurable frequency (tone ph), or to an experienced, aural and phenomenal entity (tone φ). To complete the paradigm, "tone" may also exist in a tactile mode, when the performer transforms the score into a gestural language of musculatory movements and touches. We might call this case tone g , with g meaning gesture (according to one French school, piano playing is a science of gestures) . This classification reveals that, in some of its modes of existence, music comes very close to nonverbal sign systems, not only to verbal ones (which appeal to that faded metaphor, "the language of music"). Thomas A. Sebeok has suggested that we call the basic units of all nonverbal sign systems "g-Signs" (1976: 161–62). In music one might also speak of "technemes" (Baudrillard 1968: 12), since music involves technical, concrete realization.
Because musical reality manifests itself in various ways, we have to ask what the relation is between expression and content (signifier and signified) in each mode (tactile, visual, physical, phenomenal) and whether the different musical modes can be translated into each other, thus making possible the continuity of the musical process. The first "translation" occurs in the composer's mind, with the transformation of his or her musical idea into visual notation. Next, the performer translates the score into gestural language and body techniques. Then the listener translates sound phenomena into the "language" of inner experience Finally, the most radical translation is made by those who set out to describe in words some of these modes of expression.
In the next two sections I shall survey what might be called "semiotical" ways of speaking about music. This does not imply an unambiguous solution based on safe application of an established method. Musical semiotics is a discipline in flux, a science under construction. The theories and methods of musical semiotics can nevertheless be roughly classified into two main types. On the one hand are theories proceeding from surface level toward deep structures. I call this method "structuralist," in that it represents reductionism, a reduction of sensory reality to a small number of categories. On the other hand are methods that are basically antireductionist, whose proponents argue that one should not reduce music to abstract categories that function outside the musical process, but that the significance of music is "iconic," based strictly on itself. This view emphasizes the gestural nature of music as something sensory, as a process.
Structuralist Approaches to Music
Besides the search for deep levels, the structuralist method is characterized by study of the smallest significant units of a sign system. Especially during the first phase of musical semiotics, in the 1960s, semiotics was dominated by structuralism and the direct borrowing of linguistic methods. It was thought that in music also one might distinguish the units of the first articulation (meaningful items, musical "words") and the second articulation (musical "phonemes," meaningless items). Through a sort of ars combinatoria, musical semioticians tried to build units of signification from these small atoms. All sign systems, even music, were assumed to operate like language. In this sense we may understand the argument of Lévi-Strauss in his introduction to his Mythologiques (1964: 26–30), namely, that nonfigurative painting and atonal music are not languages since they lack the level of the first articulation — meaningful and recognizable figures in painting, tonal functions in music. Accordingly, Lévi-Strauss's critique of avant-garde art has to be seen in relation to problems inherent in semiotics of the 1960s, when even composers like Pierre Boulez (1971) and Luciano Berio (1974) became interested in linguistics and structuralism. A great number of structuralist analyses inspired by linguistics appeared at that time, among them studies by Mache (1971: 75–91), Chiarucci (1973), Boiles (1973: 81–99), Ruwet (1972), and Nattiez (1976). Later, I too applied structuralist methodology to the aesthetics of music in the Romantic era, by distinguishing a particular network of mythical semes, the smallest items of mythical meaning in music, as constituting an operational reading model of certain musical texts (Tarasti 1979).
There are also structuralist methods of music analysis that have nothing to do with linguistics. Such is the case with Heinrich Schenker's theory, which we might call a structuralism before structuralism. Schenker's theory is based on "triple articulation" since he distinguishes three levels: Vordergrund, Mittelgrund, and Hintergrund (1956; see also 1954). The first of these represents the audible musical surface, while the last level is that to which the surface can be reduced. The phases through which this reduction occurs are called Mittelgrund. According to Schenker there exists only one deep structure, the Ursatz, which he claims is based on the overtone series of nature. The natural (major) triad, or "Chord of Nature," manifests itself contrapuntally; it consists of an upper, descending part , called the Urlinie, and the basic harmonic progression, called the Grundbrechung.
Schenker's theory, with its most passionate defenders and representatives in the United States, has been developed further as well as criticized. Eugene Narmour, among others, has called attention to the fact that Schenkerians always proceed from an axiom, from a preestablished structure, and everything in a musical work which is incompatible with that structure is excluded (1977: 83–86, 167–88). Thus Schenkerian analysis is restricted to langue but is totally unable to explain the changes of musical langue in a given period because new paroles, — new "idiolects" or personal styles — emerge which start to realize elements that are implicated by the Ursatz but are left out in concrete manifestations of langue. Nevertheless, Schenker's theory, like other structuralist inventions, retains its heuristic value. For example, Schenker's attempt to define the Ursatz is an effort to explain the coherence of a musical work. Musical events that take place in a work at distant temporal intervals can be understood as belonging together because they form starting and ending points of an extensive Urlinie development which Schenker called "composing out" (Auskomponierung).
What Schenker tried to make explicit with his concept of Ursatz was later paralleled by A. J. Greimas's borrowing of the term "isotopy" from physics. Isotopy designates a set of semantic categories whose redundancy guarantees the coherence and analyzability of any text or sign complex (Greimas 1980: 86–87, 112–20). In music, isotopies mean the principles that articulate musical discourse into coherent sections. If Schenker assumes that there exists only one isotopy for music, the Ursatz, we might argue that, on the contrary, there are in fact several isotopies on different levels. Further, several overlapping isotopies can function simultaneously, in which case we speak of complex isotopies. For example, two rhythmic isotopies overlap in t he Menuetto of J. S. Bach's Partita in G major (Ex. 1.1), where the rhythmic groups can be perceived as either 2 x 3 or 3 x 2, terminating in a triple rhythm at the end of each phrase. Or one may speak of harmonically complex isotopies, as in the Todesmotive of Wagner's Valkyrie (Ex. 1.2). The first chord, D–F–A, is not really a D minor triad, since F should be interpreted enharmonically as E# and the other tones as upper and lower neighboring tones to a seventh chord on C#.
Complex isotopies are also exemplified by bitonal and polyrhythmic passages, and by even larger formal units, for example, the so-called telescoping technique in sonata: A particular passage belongs at the same time to two different formal sections or isotopies; what is involved is a section that leads from development into repetition. This can be illustrated by the sections between the development and recapitulation in the first movements of Beethoven's Piano Sonatas in C major, Op. 2 no. 3, and G major, Op. 14 no. 2. Complex isotopies are also used in modern music as a deliberate stylistic device, as in Charles Ives's symphonies.
In sum, the concept of isotopy is one of the most fruitful structuralist ideas, and it remains valuable for semiotical analysis of music. Of all the semiotical and structural approaches in France in the 1960s and 70s, Greimasian theory was the only one that preserved the structuralist heritage of the isotopy concept; Greimasian theory also renewed it, even to the extent that references to the "third semiotic revolution" seem fully justified. The concept of isotopy in music will be defined in more detail before we proceed to iconic analyses of music.
Mere analysis of the musical signifier can never explain the factors that hold together musical form and guarantee its coherence. Music as such may be fragmentary; for example, long pauses may separate different sections of a musical text. Yet these sections are experienced as belonging together. Rudolph Réti called this factor the "thematic process." In his opinion there are two form-creating forces in music. One is external and is based on the segmentation and grouping of units on the manifest level of music; the other is internal and covers so-called thematic phenomena in music (Réti 1951: 138). The dramatic course, the "plot," of a work can be conceived only on this latter level. What RetJ understands by the thematic level can be redefined in terms of Greimas's concept of isotopy, producing at least five interpretations of what we mean by musical isotopy:
1. Isotopy can be a more or less achronic and abstract deep structure, such as the Ursatz in Schenker or the semiotic square in Greimas. As we relate music to such deep structures, we experience it as "making sense" (Lidov 1980). Of interest is not the mere fact that this type of structure looms in the background, but the way it manifests itself temporally in the course of a musical work. For example, given a semiotic square as in Fig. 1.1, we are interested in what order and exactly how its terms emerge during the work. What is the direction and what are the dynamics of the thematic structuring of a composition? Do we proceed from s1 to s2 and further to non-s2 and to non-s1? Or do we start with non-s2 and only at the end come to the positive term of s1? What are the dynamics of this unfolding? Is the tensional quality hesitating, expiring, retarding, accelerating, beginning, ending? These questions cannot be answered without consicteration of music's temporal unfolding.
According to Greimas (1979) , there are three temporal categories: inchoativity, durativity, terminativity. Also, his category of perfectivity/ imperfectivity describes the relation of the temporal process to the initial problem of the composition. The "problem" does not necessarily refer to any initial "lack" in a narrative, or "conflict" in a mythical sense. Rather, what is involved is that every tone heard after the first puts the balance of the work in question and creates a demand for the return of that balance; it would be difficult to imagine a musical universe lacking all tensions caused by durativity. In chapter 2 the fundamental problems of temporality in music \Viii be examined in the light of semiotics.
2. What Réti meant by thematicity can form an isotopy of music. For example, Beethoven's Piano Sonata in E major, Op. 109, consists of three movements whose interpretation is problematic, since they break the normal alternation of fast and slow movements in a sonata. First comes a movement where vivace and adagio alternate, then a fast middle movement in prestissimo, and at the end a large variation section built on a lied-type theme ( Ex. 1.3). The structure of the sonata becomes clear when we notice that the whole musical substance of the first two movements has been inferred from the lied theme of the last movement. All that is heard in the work before the emergence of this theme belongs to the "not yet."
Excerpted from A Theory of Musical Semiotics by Eero Tarasti. Copyright © 1994 Eero Tarasti. Excerpted by permission of Indiana University Press.
All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.
Excerpts are provided by Dial-A-Book Inc. solely for the personal use of visitors to this web site.