The Analysis and Cognition of Melodic Complexity: The Implication-Realization Model.
In my review of the previous volume (Notes 49 [1992-93]: 588-90) I emphasized how Narmour's theory of melody assumes an interplay between listeners' innate responses and their learned, culturally specific expectations. The review itself, in fact, provided a typographical example of how dynamic that interplay can be. Through the inadvertant omission of a line of text, the sentence "He takes the Gestalt principle of implication governed by similarity and posits a symmetrical concept of implication governed by differentiation" made its way into print as "He takes the Gestalt principle of implication governed by differentiation." The innate, "bottom-up" response to the matching of a string of words ("of implication governed by") overrode any learned, "top-down" knowledge of syntax and rhetoric. With respect to melody, as Narmour states in his new volume,
The invocation of top-down structures depends on the nature of the bottom-up input that activates them. Without identifying low-level implications and realizations, higher-level relations do not cognitively come into play. Without them, no voice leading emerges. Without them, no counterpoint exists. Without them, there is no harmonic motion. Despite thousands of analyses to the contrary, listeners do not attend to music just in order to assimilate the bottom-up foreground to higher-level, top-down relations of voice-leading structure and key. Indeed, it is the dynamic, nonreductive, individuated melodic motion at the note-to-note level that captivates music lovers.
The desire to avoid a melodic analysis that is "crudely reductive" forced Narmour to confront the question of whether shades of gray ought not to be allowed into the mostly black-and-white world of music theory. Take the case of structural tones. Implicit in most tonal analyses is the notion that of each tone it should be asked, "Are you structural or ornamental?" Little tolerance has traditionally been shown to those tones that diffidently respond, "A little bit of both." Narmour concludes, by contrast, that in melodic analyses "A little bit of both" is a legitimate response. He recognizes, between the extremes of the completely ornamental and the completely structural, two intermediate shades: "articulation" and "formation." "Articulation . . . is a weak degree of closure wholly contained on the level of its occurrence, whereas formation . . . constitutes a stronger degree of closure, one that comes close to producing a new hierarchical level". Viewed in terms of Narmour's theory as a whole, his global scale of degrees of structuring is a logical outgrowth of the separate scaled parameters that go into each analysis; that is, the shades of closure and nonclosure calculated for component durations, intervals, dissonances, and so forth, yield, in the aggregate, varying shades of overall closure.
Just as important as determining which tones serve to structure the beginnings and endings of perceived patterns is the determination of which tones do not. The implication-realization model generates far more implications than realizations. So it is crucial for it to specify why many implications remain unrealized and thus why many tones do not give rise to higher-level structures. The earlier volume treated instances that can be explained largely with reference to a single parameter. For example, extremely brief tones, extremely dissonant tones, tones too distant in time from their implicative antecedents, or tones falling on an extremely weak part of the meter rarely create higher-level structures. This volume focuses on the more interesting and more difficult instances where several parameters must be considered jointly to determine why a tone remains nonclosural.
To aid readers in determining when small melodic archetypes are perceptually likely to combine, Narmour provides eight rules of thumb. The first, for example, specifies that closure is deferred by "the occurrence of strong nonchordal dissonance on a metric accent in additive or weakly cumulative durations (provided the resolution itself is weak)". This means, for instance, that the first tone of a 4-3 appoggiatura will remain nonclosural (= nonstructural) even if its falls on a downbeat, unless it happens to be accented by having a significantly longer duration than the tone that precedes it. The common-sense nature of such a prescription masks the surprising inferences that result from its strict application. Thus cadences by J. S. Bach and Richard Wagner, which because of shared harmonic, contrapuntal, and metric features might typically be viewed as exemplars of a common tradition, could, given Narmour's evaluation of their featural interactions, be analyzed as having radically different perceived structures. Thus a distinguishing characteristic of Narmour's analysis is the degree to which salient temporal aspects of music--duration, relative duration, and meter--are taken fully into account.
The 435 melodic excerpts presented in this volume, coupled with the hundreds contained in the previous volume, constitute a huge analytical corpus that will challenge the field for some time to come.
ROBERT O. GJERDINGEN Center for Advanced Study in the Behavioral Sciences, Stanford
|Printer friendly Cite/link Email Feedback|
|Author:||Gjerdingen, Robert O.|
|Article Type:||Book Review|
|Date:||Jun 1, 1994|
|Previous Article:||Guidelines for Style Analysis, 2d ed.|
|Next Article:||Opera Batak, das Wandertheater der Toba-Batak in Nord-Sumatra: Schauspiele zur Wahrung kultureller Identitat im nationalen Kontext, vols.|