Patterns and Strategies: Four Perspectives of Musical Characterization.

Abstract

We cannot say that an analysis of a piece of music is merely the discovery of structure hidden in that piece by its composer. The patterns and features that an analyst points to are, for the most part, just those that interest him, and st and ards of interest are subjective and flexible. No single analysis of a piece, no matter how elaborate, comprehensive, or persuasive, should ever be regarded as an absolute, all-embracing analysis. This paper examines four of the many different viewpoints to be recognized in characterizing analyses. Each viewpoint suggests a different perspective of analysis and analytical strategies; each leads to different answers to the questions "What kind of analysis is this?" and "What are its uses?" (I) "Is this analysis true? Is it accurate?" Whether an analysis is powerful or pedantic, informative or trivial, has little to do with its accuracy. Nonetheless truth is a critical factor in characterizing the individual sentences of an analysis, as the work of philosophers such as W. V. Quine and Nelson Goodman suggests. Most critically, there is no sharp distinction between the truth conditions of observational and theoretical sentences; what an analyst observes in a piece depends on his training and expectations, that is, on the theoretical sentences he already believes. An obvious way of characterizing individual words and expressions used in an analysis is to say what they are true of (refer to). However, we can also characterize expressions in other fruitful ways: for example, as monadic or relational, as relatively literal or metaphorical, and as relatively determinable or undeterminable. (II) "How systematic is this analysis? How sophisticated is it?" Segments of a piece are transformed into prolongations of things (typically of chords or motives) when names chosen from some implicit or explicit vocabulary are applied to them. We pay special attention to certain successions of prolongations in an analysis, namely the progressions; our sense that a piece is directed or that its events are functional depends on our hearing in terms of progressions of one kind or another. A vocabulary for prolongations along with a system of functional categories and progressions is a syntax for any piece that it is applied to. The extent to which any analysis can be translated into such a general language of prolongations and progressions is one measure of its sophistication. (For example, Heinrich Schenker's analyses are almost completely accessible to such translation.) Analytical sketches can similarly be formulated in a general but rigorous notation for prolongations and progressions; the most elaborate examples in this paper are sketches of portions of the Aria from Bach's "Goldberg" Variations, Brahms's Intermezzo in E (Op. 116/6), Chopin's Etude in C (Op. 10/1), and Webern's orchestra piece (Op. 10/4). (III) "How appropriate is this analysis to this piece?" (IV) "How interesting is this analysis?" An analysis is appropriate only if it reflects the application of a relatively familiar syntax; deciding upon syntaxes that are appropriate to particular pieces involves the hypothesis of theories of musical coherence. Whereas appropriateness (familiarity) is a relatively conservative criterion, the interest of an analysis is associated both with its novelty and with our concern for inductive, inferential thought. There is a dynamic interaction between our intuitions of appropriateness and interest, an interaction which is responsible for much of what is compelling and creative in musical thought.