Music Theory Online

The Online Journal of the Society for Music Theory

[Contents of vol. 3.2] [References] [MTO Home] [SMT Home] [MTO talk]

Volume 3.2:

Latham, Edward D.*

Review of Ethan Haimo's Article, "Atonality, Analysis, and the Intentional Fallacy,"Music Theory Spectrum 18.2 (Fall 1996), 167-199

KEYWORDS: Haimo, Schoenberg, Forte, atonality, intentionality, set theory



Ethan Haimo's article, "Atonality, Analysis and the Intentional Fallacy," provokes thought on three issues: 1) the issue of conscious vs. unconscious intentions; 2) the issue of documentary evidence and its uncritical acceptance; and 3) the evolution of pitch-class set theory in the last twenty-five years. The present author addresses these three issues and challenges Haimo's conclusion that analyses that do not take intentions into account effectively replace the composer with the theorist as an authority figure.

[1] Ethan Haimo's recent article, "Atonality, Analysis, and the Intentional Fallacy," is both interesting and provocative.{1} If read from a certain point of view, it would seem that Haimo is interested in resurrecting Nattiez's tripartition (poietic/neutral/esthesic) and paradigmatic method of analysis, discussing as he does the notions of "coherence based on repetition," "the score itself," and "two types of statements about music," involving composition (poietics) and perception (esthesics). But such is not the case, and the tripartition will have to wait for another time. What Haimo is interested in, however, is the issue of intentionality and its putative role in musical analysis, and Nattiez's name will therefore resurface in this review in connection with this most thorny, but potentially revelatory of music-analytical topics.

[2] Haimo divides his article into four sections, which might be labeled "Schoenberg and Intentionality," "Intentionality Revisited," "Forte and Intentionality," and "Schoenberg and Developing Variation." In his first section, he presents the case against Schoenberg's intentional use of pitch-class sets (as they are conceived of in Allen Forte's system) in his atonal compositions. After presenting a tidy, if not entirely objective, encapsulation of the controversy that Forte's undaunted application of his theory to the analysis of that very same music has engendered, Haimo produces several handsome plates of very official-looking sketch materials. His approach, like Perle's approach to Berg before him, is to infer, based on the superabundance of row charts and other pre-compositional calculations in the sketches for the twelve-tone works, that Schoenberg did not compose with pitch-class sets in his atonal music, since no similar materials have yet been discovered for those works.

[3] Haimo's argument here consists of a very carefully hedged bet. In his second section, where he again provides a helpful summary of intellectual controversy, this time concerning the intentional fallacy itself, Haimo cautiously draws the distinction between "conscious intentions" and "intuitive behavior," declaring himself to be interested exclusively in the former (p. 178, n. 29). In making such a declaration, however, he is in fact dodging a good deal of the issue altogether. If "documentary evidence" from a composer like Wagner, who claimed to have "forgotten all theory" in the composition of his masterworks, is to be accepted, then unintentional, or intuitive, use of structural devices must be included in the consideration of intentionality (see par [6] below). Haimo, however, seems to believe categorically that "complicated and thoroughgoing procedures" cannot possibly be intuitive (p. 179).

[4] As he continues his discussion of intentionality, Haimo isolates two possible types of music-analytical statements, the first concerning the compositional process and the second concerning the score itself. The first of these, he maintains, must be validated by "documentary or stylistic support," while the second requires only perceptual confirmation. He also introduces the notion of persuasiveness, yet implicitly exempts type-one statements from its requirements. It seems reasonable, however, to apply the requirement of persuasiveness (or perhaps "meta-persuasiveness") to both types of statements, in that any analytical essay, whether type-one or type-two, is an attempt to persuade a reader to adopt a particular point of view.

[5] One is struck by the seeming naivete of Haimo's trust in "documentary evidence." Not only does he seem to consider it a "get-out-of-jail-free card" with respect to persuasiveness, but he includes the widest possible assortment of documents under the rubric of "evidence," including letters, theoretical works, notes of explanation, explanatory essays, and testimony by students or colleagues, in addition to the aforementioned calculations or diagrams found in sketch materials (p. 179). As a composer, Haimo must be aware that his colleagues are not always to be trusted on the subject of their own works; one need only think of the numerous anecdotes concerning Stravinskian and Schoenbergian smokescreening on the subject of the influence of the text on their compositional decisions in texted works. For a more specific example of this unreliability, one could look to Jonathan Bernard's critique of the use of documentary evidence by Nattiez in his analysis of Density 21.5. {2} In a recent review of a Spectrum article published in Music Theory Online, Jane Clendinning summarizes the problem, noting that "a composer's comments on his music should be examined critically (comments may be used to obfuscate rather than enlighten)."{3}

[6] In a posting to the smt-list that ignited a brief but lively exchange of ideas, I discussed intentionality and the uncritical acceptance of documentary evidence at some length, in reference to the works of Wagner and Webern.{4} Concerning the former, my conclusion was that Wagner claimed he was disregarding his "theoretical scruples," as he put it, both to distinguish himself as a creative individual and to throw music-analytical sleuths off his trail. By saying that he had composed Tristan in "complete freedom . . . in an uncritical frame of mind," Wagner was attempting to erase his tracks so that no one could follow in his footsteps, thereby preserving an aura of mystique about the compositional process. The lesson to be learned here is that composers' comments on their own music are not dependable as sources of documentary evidence.

[7] By the time he arrives at his critique of Forte's analysis of op. 11/1, Haimo has worked himself into a pretty good lather. Although his critique of Forte's segmentational decisions is perceptive and valid, several elements of his argument weaken the whole. First of all, he uses the earlier of Forte's two analyses of op. 11/1 (from "Magical Kaleidoscope" (1981)), which, he admits in a footnote (n. 39), contains much less detail and analytical information than the later one, from "Pitch- Class Set Analysis Today" (1985). Secondly, his discussion of Forte's segmentational criteria is based on an outdated article ("Sets and Nonsets in Schoenberg's Atonal Music" (1972)). The fourth criterion that Haimo cites from the article (that the set ought to be "an 'atonal' set, not a set that would occur in a tonal work," n. 42) reflects Forte's early, pre-genera, thinking on the subject of segmentation; "tonal" sets such as 4-27 and 4-28 are regularly featured in his current analyses of atonal repertoire.

[8] The use of the earlier analysis is indicative of Haimo's almost exclusive reliance on sources dating from the ten-year period from 1972-1981 in his critique of Forte's methods, with the majority of his theoretical citations drawn from "Sets and Non-Sets in Schoenberg's Music" (1972), and The Structure of Atonal Music (1973).{5} He draws the rest of his citations from works written between 1977 and 1985, thereby excluding many important works, including the articles on Berg (1985, 1991), Debussy (1989), Musorgsky (1990), Liszt (1990), and Webern (1994). Moreover, many of his criticisms of Forte are based on citations drawn from the polemical exchange between Forte and Taruskin (e.g. "Letter to the Editor from Allen Forte"), rather than from the more detailed and comprehensive expositions of pitch-class set theory to be found elsewhere.

[9] By contrast, Haimo conscientiously utilizes the most up-to-date information available in support of his own views, relying upon articles published between 1990- 1994.{6} In the twenty-fifth anniversary year of the publication of "Sets and Non-Sets," (and on the eve of Structure of Atonal Music's 25th birthday, as well) it seems appropriate, indeed necessary, to take into account the evolution of Forte's own thinking on the subject of pitch-class set analysis, as well as the contributions of his students and colleagues on the subject.

[10] In general, the sources Haimo omits point to a new method of segmentation, that, in 1981 when "The Magical Kaleidoscope" was written, Forte had not even begun to develop--namely, segmentation based on a linear approach to the analysis of post-tonal music. After immersing himself in Schenker's theories in order to complete his Introduction to Schenkerian Analysis with Steven Gilbert in 1982, Forte published three articles devoted to linear analysis of late nineteenth-century music: one on the Brahms Alto Rhapsody (1983), one on the Brahms C-minor String Quartet (1983), and one on Mahler's Fifth Symphony (1984). These articles, cited in Beach (1985) as an example of how to extend the possibilities of Schenkerian analytical techniques, began to develop a system of motivic analysis at multiple structural levels that has been a focus of Forte's work ever since.

[11] "Pitch-Class Set Analysis Today" (1985), then--the most recent article to which Haimo makes reference--is actually Forte's own attempt to signal the end of an initial phase in the evolution of pitch-class set theory and the beginning of a new phase in which a more sophisticated analytical method might be developed. He presented a general outline of such a method in "New Approaches to Linear Analysis" (1988). Analytical decisions, he says, are to be made on the basis of three criteria: 1) aural salience; 2) motivic replication; and 3) coincidence of onset and closure.{7} These criteria are used to identify primary and secondary segments of a different sort: linear and harmonic motives that are present at different levels of musical structure. This method leans far more heavily toward Haimo's "type-one" assertions about music (statements that explain something about how the music itself sounds), relying as it does upon aurally salient criteria like register, dynamics, timbre, texture, and rhythm.

[12] That "New Approaches to Linear Analysis" was published in 1988 is surely more than coincidental; it is a direct response to the criticisms of the work of Salzer (1952), Travis (1969), and Morgan (1976) by Baker (1983) and Straus (1987). Forte attempts to distance himself from Schenker and the problems associated with invoking his ideas for atonal music by using the general term "linear analysis," rather than "Schenkerian analysis." Perhaps, if only to rankle Eugene Narmour, it would be best to dub this method "post-Schenkerian" linear analysis, since it is clear (and Forte makes no effort to deny it) that his ideas are inspired to a large extent by the work of Schenker.

[13] Forte has applied linear analysis in a series of important articles, including the pieces mentioned in paragraph [8]. Most recently, he has turned again to more traditional Schenkerian techniques to explain the music of Porter and Gershwin, among others (inThe American Popular Ballad of the Golden Era 1924-1950 (1995)). Perhaps most relevant to Haimo's critique, however, is Forte's discussion ofDas Buch den Hangenden Garten in "Concepts of Linearity in Schoenberg's Atonal Music: A Study of the Opus 15 Song Cycle" (1992). Given that Haimo is specifically interested in discrediting Forte's approach to Schoenberg's atonal music, it seems rather surprising (and unfortunate) that he chooses to ignore this particular article.

[14] If Forte's linear approach represents a fundamental change in methods of segmentation and derivation of structural sets, then "Pitch-Class Set Genera and the Origin of the Modern Harmonic Species" (1988) represents a change in methods of categorizing the resultant sets. Intended to supercede the K and Kh relations as a means of describing the relationships among the different sets employed (or discovered,pace Haimo) in a composition, the system of genera is clearly a further refinement of pitch-class set theory, and as such it, too, needs to be included in any evaluation of the theory and its analytical application.

[15] As far as Haimo's own analytical proposals are concerned, they are built on tenuous conclusions about Schoenberg's relationship to the idea of developing variation. Whether he actually used developing variation as a compositional tool or not, Schoenberg clearly had an agenda for citing it as an important influence on his work. As Haimo notes, Schoenberg discusses the concept in an essay on Bach and it has since been linked to Brahms, as well (p. 192, n. 55). As a neoclassicist and a great admirer of Brahms, Schoenberg would have wanted to be associated with the same compositional technique.{8} Furthermore, Schoenberg's definition of coherence as being based on repetition, and his association of coherence with comprehensibility and the need for effective communication (p. 192), points to a composer attempting belatedly to justify some very complex music. Similarly, in his analysis of op. 22, Schoenberg is trying to prove the coherence of his music, so that he will be less "difficult to understand" (p. 193). Thus, when he declares that his analytical approach "may well be compatible with what we know of Schoenberg's intentions," (p. 198) Haimo is begging the question: what exactly do we know about them?

[16] Further problems arise when Haimo attempts to appropriate Schoenberg's notion of developing variation for the purposes of his own analysis. Although Schoenberg's general definition of the concept refers to a procedure of variation in which the changes proceed toward the goal of allowing new ideas to arise (p. 192, emphasis added), Haimo prefers to focus on information gleaned from Schoenberg's op. 22 analysis to justify his identification of motives that "come back and fill in the missing steps" from one variation to another, more remote one (p. 195). Yet, in his summary of developing variation in op. 11/1, Haimo again refers to a process in which "changes made in one place invariably are taken up later to create new ideas, which are themselves subjected to further transformations in a continuous process of forward motion" (p. 197, emphasis added). Thus, Schoenberg is invoked twice to justify differing analytical points.

[17] And what of Haimo the composer-theorist's intentions? From his conclusion, it would seem that he fears "the substitution of the analyst for the composer as the figure of authority for the validity of interpretations" (p. 199). Yet he argues (persuasively) for the need for persuasiveness in type-two statements about music (p. 179), thereby implying that the reader (or audience member) towards which an interpretation is directed is the ultimate arbiter of validity, and I strongly support this view. His concern that the composer will be reduced to "an almost accidental figure" seems to stem from his compositional side; yet his theoretical side ought to tell him that "the analyst's assumptions, and not the composer's" are unavoidably the determining features of an analysis, and that "flights of analytical fancy" are desirable, not suspect. The task is to convince the reader to come along for the ride.

Edward D. Latham
Yale University
Department of Music
143 Elm St.
New Haven, CT 06511


[Contents of vol. 3.2] [References] [MTO Home] [SMT Home] [MTO talk]

* Return to beginning of article


1. Ethan Haimo, "Atonality, Analysis, and the Intentional Fallacy," Music Theory Spectrum 18.2 (Fall 1996), 167-199.
Return to text

2. Jonathan W. Bernard, "On Density 21.5: A Response to Nattiez," Music Analysis 5.2-3 (July/October 1986), 207-232.
Return to text

3. Jane Clendinning, "Review/Article of Miguel Roig- Francoli's 'Harmonic and Formal Processes in Ligeti's Net-Structure Compositions'," Music Theory Online 2.5 (July 1996), [2].
Return to text

4. Edward D. Latham, "8-28/Intentionality," smt-list communication, April 1995. See also the source communication by Stephen Soderberg on Wagner (which includes all the relevant quotes from Mein Leben), and the responses by a variety of theorists.
Return to text

5. "Sets and Nonsets" is referred to three times; The Structure of Atonal Music four times.
Return to text

6. A brief summary of these sources: Perle (1990), Haimo (1990), Auner (1991), Boss (1992), Keathley (1992), Hatch and Bernstein (1993), Neff (1993), Neff and Cross (1994).
Return to text

7. Allen Forte, "New Approaches to Linear Analysis," Journal of the American Musicological Society 41.2 (Summer 1988).
Return to text

8. Martha Hyde discusses Schoenberg's neoclassicism in the article immediately following Haimo's: "Neoclassic and Anachronistic Impulses in Twentieth-Century Music," Music Theory Spectrum 18.2 (Fall 1996), 200-235.
Return to text

* Return to Beginning

[Contents of vol.3.2] [References] [MTO Home] [SMT Home] [MTO talk]

Copyright Statement

Copyright � 1997 by the Society for Music Theory.
All rights reserved.

[1] Copyrights for individual items published in Music Theory Online (MTO) are held by their authors. Items appearing in MTO may be saved and stored in electronic or paper form, and may be shared among individuals for purposes of scholarly research or discussion, but may not be republished in any form, electronic or print, without prior, written permission from the author(s), and advance notification of the editors of MTO.

[2] Any redistributed form of items published in MTO must include the following information in a form appropriate to the medium in which the items are to appear:

This item appeared in Music Theory Online
It was authored by [FULL NAME, EMAIL ADDRESS],
with whose written permission it is reprinted
[3] Libraries may archive issues of MTO in electronic or paper form for public access so long as each issue is stored in its entirety, and no access fee is charged. Exceptions to these requirements must be approved in writing by the editors of MTO, who will act in accordance with the decisions of the Society for Music Theory.

This document and all portions thereof are protected by U.S. and international copyright laws. Material contained herein may be copied and/or distributed for research purposes only.

prepared by
Nicholas S. Blanchard, MTO