markus heuger's
beabliography

Beabliography

Abstract 0105

Heinonen, Yrjö (1996), "Semantic and episodic models in music. A cognitive approach to the analysis of compositional influences." Paper presented at the Joint International Conference of Systematic Musicology and Cognitive Musicology, 8-11 September 1996, Brugge (Belgium).

A procedure for analysing compositional influences (models) is proposed. Theoretically the procedure is based on the notions of semantic and episodic memory (Tulving), explicit and implicit memory (Nelson & Schreiber & McEvoy), and the transfer effect. The methodological basis of the procedure is based on the following criteria (Hermerén): (1) the temporal requirement, (2) the requirement of contact, (3) the requirement of similarity, and (4) the requirement of change. The procedure was tested in two case studies: "Michelle" by John Lennon & Paul McCartney and the "Allegro Appassionato" theme of the A Minor Quartet (op. 132) by Ludwig van Beethoven. Biographical data was surveyed in order to meet the first two requirements, while musical analysis was used in order to meet the others. In both cases the semantic models of the test composition were common in the comparison material, although explicit semantic models were found only in the case of "Michelle". A few explicit and implicit episodic models were found in both cases. It proved difficult to distinguish transfer models from implicit models. Some evidence was, however, found to support the conclusion that the compositions belonging to the same genre were also the most probable source of transfer.

Summary

1. Introduction

The aim of the paper is to present a procedure for the analysis of the relationship between an assumed compositional model and a new composition based on this model. The main problem is the degree to which certain similarities found between two or more compositions regarded are regarded as indirect (in other words, as a result of a common style) and the degree to which it is reasonable to assume a direct influence between two compositions. The procedure was tested in two case studies representing different musical traditions. The first study is on the compositional models of "Michelle" by John Lennon and Paul McCartney. A more detailed version of this study can be found (unfortunately only in Finnish) in Heinonen 1992. The second study on the compositional models of the "Allegro Appassionato" theme of the A minor Quartet (Op. 132) by Ludwig van Beethoven. This study, too, will be reported in a more detailed form in the future.

1.1 Theoretical foundations

A compositional model is regarded here as a more or less abstract conception of existing music with an emphasis of the features essential to the composer or the compostional task in question (cf. Laske 1989; Heinonen 1995). The models derived from a number of existing compositions through a process of generalization are called semantic, while those based on a single composition are called episodic (cf. Tulving 1983). Further, the influence of certain episodic models is rather indirect than direct: they undoubtedly change the semantic models of the composer before the compositional process of the new composition. The composer, however uses these new semantic models rather than original episodic models during the compositional process. This kind of model is assumed to be based on transfer effect and is called the transfer model. Finally, a composer may or may not be aware of using a certain (aspect of a certain) composition as a model of a new composition. If he/she is aware of using a certain model, this model is said to be explicit, if not, the model is said to be implicit (cf. Nelson & Schreiber & McEvoy 1992).

1.2 Methodological foundations

If the model is assumed to be episodic (based on a single composition with respect to a certain feature), the following requirements must be fulfilled if one wishes to show that the assumed model has actually influenced the compositional process of the new composition (cf. Hermerén 1975; Platoff 1988): (1) the model has to be composed before the new composition (the temporal requirement), (2) the composer has to have heard — or seen — the model (the requirement of contact), (3) there has to be noticeable similarity between the model and the modelled (the requirement of similarity), and (4) the contact with the model is assumed to change the further output of the composer (the requirement of change). The first step is to restrict the comparison material in such a way that it comprises an adequate sample of compositions belonging to the style(s) in question and at the same time takes the temporal requirement into account. The second step is to find, by means of music analysis, the songs that meet the requirements of similarity and change. Similarities are compared on several levels, which may roughly be reduced as surface and deep structures. Surface level models are called scripts, while deep level models are called plans (cf. Schank & Abelson 1977; Gjerdingen 1988). The third and last step is to test, by means of biographical study, which songs meet also the requirement of contact.

2. Study I: The Compositional Models of "Michelle" by Lennon & McCartney

2.1 Procedure and Materials

Some 2500 popular compositions (consisting mainly of jazz, popular songs and rock) from the 1920's to 1965 were examined in order to find the songs that meet the requirements of similarity and change. To be more specific, the task was to find the features "Michelle" shares with other compositions of the sample, especially with regard to its overall form and tonal plan, its period and phrase structure, and its melodic and harmonic schemata. In order to meet the requirement of contact standard biographies, interviews, press conference statements, and descriptions of the recording sessions were examined, especially in order to find any information concerning the chronology, contacts with assumed models, explicit statements about possible influences and so on.

2.2 Results

The formal structure of "Michelle" may be characterized as follows: the song consists of an intro, an AABA-structure as the basic formal unit of the song, reprises of the A and B sections, a guitar solo in the middle of these repetitions, and a coda ending with the reprise of the guitar solo. This structure is — with only slight variations — probably the commonest one in the pop-rock tradition from the late 1950's through the middle of the 1960's.

The tune of the A section is based at the background level on a descending linear progression in so called gypsy minor, interrupted on the 2nd degree (5 --- 4#-3-2 //). The corresponding harmony is based on the following chords: I --- vii7/V. At the foreground the melody proceeds by a descending sequence of a motif based on upward and downward fourths, harmonized by a chord progression based on the circle of fifths (I-iv7-VII). Both the background and the foreground models are quite common in the Tin Pan Alley tradition (in other words, in American popular songs in the first half of the 20th Century). The opening of the B section of the melody is based on a 8-9-8 script, which is harmonized by a tonic minor chord. This script (which also may be harmonized by iv or VI) is quite common in popular songs written in a minor mode. In the doo wop and teen rock tradition of the late 1950's and early 1960's it is very common as a major mode version (6-7-6, usually harmonized by IV or vi). The B section is closed, again, by a descending linear progression 5 --- 4-3-2 //. However, here the fourth is natural, not sharp as in the case of the A section. Also the chord progression (i --- VI7/V) at the background level is very similar to the tonal plan of the A section. At the foreground the chords are planned according a descending chromatic melody line from the tonic to the dominant, resulting the following chord progression: Fm, C+, Fm7, Fm6, Db, and C. Also this foreground pattern is very common in the Tin Pan Alley tradition.

Lennon and McCartney were obviously quite aware of the general stylistical models of "Michelle". For McCartney, "Michelle" was originally a joke French tune, and Lennon described it as a strict ballad with some bluesy edge. McCartney has also pointed out some more specific stylistical models concerning the arrangement: the acoustic guitar in "Michelle" was in fact his version of Chet Atkins' finger picking technique, and the bass line was heavily influenced by Brian Wilson's way of using inverted chords (especially the 5th inversion). These general and specific models may be regarded as explicit semantic models of "Michelle". A closer analysis of the comparison material reveals that the overall form, the harmonic-tonal plan, the rather complex period and phrase structure, and certain melodic-harmonic patterns of "Michelle" are evidently based on conceptions derived from the Tin Pan Alley tradition. On the other hand, the foreground metrical and rhytmical patterns, tone repetitions and pentatonic aspects of "Michelle" are in turn based clearly on conceptions derived from the teen-rock and doo wop tradition of the late 1950's and --to a lesser degree — on the rhythm'n'blues and rock'n'roll tradition of the 1950's and the early 1960's. This kind of general models were regarded as implicit semantic models of "Michelle".

In two cases the use of a certain song as a model of "Michelle" was confirmed by personal statements. These songs are "I Put A Spell On You" (Hawkins) and "Habanera" (Bizet). These two songs are also considered as explicit episodic models of "Michelle". Two other songs share noticeable similarities with "Michelle", appearing on several levels, and they also meet all four requirements of influence. These songs are "All I Have To Do Is Dream" (Bryant & Bryant) and "A Taste Of Honey" (Scott & Marlow). "Chim Chim Cher-ee" (Sherman & Sherman) may also be added to this category, although the songwriters' contact with it can only be inferred on the basis of circumstancial evidence. These three songs are considered as evident implicit episodic models of "Michelle". Further, in many cases two or three requirements of influence are met, but the absence of the remaining ones leaves the question of possible influence open. If the similarity is very prominent and the contact seems — on the basis of circumstantial evidence — probable, the relationship seems intuitively very convincing. Representative examples of this case are "All The Things You Are" (Kern & Hammerstein), "Les Feuilles Mortes" (Kosma & Prevent, known as "Autumn Leaves" in English) and "Samba De Uma Nota So" (Jobim & Mendonca, known as "One Note Samba" in English). Also these three songs are considered here as implicit episodic models of "Michelle".

In some cases the influence is obviously indirect (based on transfer effect): the biographical data clearly shows that contact with a certain song has essentially changed the compositional models of Lennon and McCartney, and "Michelle" is — to a certain degree — based on such models. Nonetheless the perceivable similarity is too abstract or related only to a single aspect on a single level for us to be able to assume a direct (episodic) influence between such a song and "Michelle". "Summertime" (Gershwin) and "Long Tall Sally" (Blackwell & Penniman & Johnson) may be mentioned as examples of this case. Also many Lennon-McCartney compositions written before "Michelle" may be considered as examples of this category, perhaps most notably such McCartney ballads as "And I Love Her" and "Yesterday". Some of the songs considered above as implicit semantic models may as well be regarded as transfer models (this is so because semantic models are assumed to be acquired by transfer). It may be concluded that the songs representing the same genre — the ballads written in the Tin Pan Alley style — were the most probable transfer models of "Michelle".

3. Study II: The Compositional Models of the "Allegro appassionato" theme by Ludwig van Beethoven

3.1 Procedure and materials

The thematic catalogues of Haydn, Mozart, and Beethoven himself (Hoboken, Köchel, and Kinsky-Halm, respectively) were examined in order to find which opening themes written by the three Viennese classics meet the requirements of similarity and change, as compared to the "Allegro appassionato" theme (or, to be exact, the first eight bars of the theme). The Piano Sonatas and a selection of works for chamber ensembles and/or piano by the same composers were used as additional material (this was done because the catalogues give only the incipit of each work or movement of it). In order to meet the requirement of contact standard biographies, descriptions of contemporaires, letters, conversation books and other biographical material were examined. Special attention was given to the chronology — exact dates or years --, contacts with assumed models, explicit statements about possible influences and so on.

3.2 Results

The "Allegro Appassionato" theme comprises eight bars, which can be divided further into the opening script (bars 1-2), the middle script (bars 3-5), and the closing script (bars 5-8). The theme itself is preceded by a two-bar "curtain" (introduction).

The most obvious deep level model of the theme is the overall tonal schema i-III. This is, of course, one of the three standard procedures in a minor mode, the other two being i-i and i-v. The i-III procedure is especially typical to Haydn and Beethoven while Mozart tends to favour i-v instead of i-III. The main melody is based on a changing-note schema 1-7; 4-3, which, in the "Allegro appassionato" theme, is interrupted at the 4th degree (harmonized by V/III). The opening, the middle, and the closing of the theme are based on scripts common in Viennese Classic tradition. With regard to Beethoven, the middle and the closing scripts appear almost exclusively in the works of so called Bonn and early Vienna periods. Thus it may be concluded that the use of these scripts in "Allegro appassionato" may be traced back to the days of Haydn and Mozart. In many cases the opening, the middle, and the closing scripts were immersed in a more general plan. The opening script of "Allegro appassionato" was also in some other themes used as the beginning of the i-III schema. Correspondingly, the closing script was in certain cases used as the ending of the i-III schema. Nevertheless, only in the "Allegro appassionato" theme both the opening script and the closing script are immersed in the i-III schema. It was also shown that if one of the models of the "Allegro appassionato" theme appear in a certain movement of a multi-movement work, the same model commonly appears also in other movements of the same work. Especially the opening script tends to appear more than once within a multi-movement work.

No clear statements in Beethoven's letters or conversation books concerning the models of the "Allegro appassionato" theme were found. However, it is a well established fact that the early sketches of the theme were originally planned for the Ninth Symphony (op. 125), which also shares many significant structural similarities with the A minor Quartet (both in a general and on a particular level). On this basis certain aspects of the Ninth Symphony were regarded as explicit episodic models of the "Allegro appassionato" theme. It was also shown that many other works shared noticeable similarities with the A Minor Quartet as a whole and/or with the "Allegro appassionato" theme in particular, and also met all four requirements of influence. Certain aspects of Mozart's Requiem in D minor and String Quartet in A major (K 626 and K 624, respectively), and Beethoven's own String Quartet in F major (op 18/1) were perhaps the most obvious examples of this kind of implicit episodic influence. There were, again, cases in which the similarity was very significant but contact with the assumed model was inferred only on the basis of circumstantial evidence. The String Quartet in D minor and German Dance in A major by Mozart (K 421 and K 567/5, respectively), and the Piano Sonata in C# minor by Haydn (Hob XVI/36) were representative examples of this category. Also certain aspects of these works were considered as implicit episodic models of the "Allegro appassionato" theme.

Of course, many other compositions by Haydn, Mozart, and Beethoven himself meet — to a greater or lesser degree — all four requirements of influence (it is common knowledge that Beethoven was well acquainted with the music of Haydn and Mozart, although it is as obvious that he did not know all the works written by them). In this sense it proved very difficult — if not impossible — to draw a line between implicit semantic models and transfer models. It seems, however, reasonable to assume that the works of the same genre — in this case the string quartets (especially Haydn's op. 33, Mozart's "Haydn Quartets", and Beethoven's own op. 18) — were, again, the most probable transfer models.

4 Conclusion

A procedure for the analysis of compositional influences was proposed and tested in two case studies representing different musical traditions (Viennese classicism and rock music of the 1960s). In both cases it was shown that the main semantic models of the test composition were quite common in the comparison material. However, explicit semantic models were found only in one case (Lennon-McCartney). 1-2 explicit episodic models and 5-6 implicit episodic models were found, again, in both cases (this does not mean, however, that the episodic models of the test compositions are restricted only to these 6-8 models found in this study). Although it proved quite difficult to distinguish transfer models from implicit semantic models on the one hand, and implicit episodic models on the other, it was concluded that the compositions of the same genre were the most probable — although by no means the only — source for this kind of influence. This holds, again, to both cases examined in this study. The search for "similarities" or "influences" is by no means new in musicology (cf., for example, LaRue 1961; Meyer 1973). This kind of study is sometimes called "Reminiszenzjakt" or "tone-detecting". However, as far as I know, it has never been based at the same time on a firm psychological theory and sufficiently developed musicological methods.

References

Gjerdingen, R. O. 1988. A Classic Turn of Phrase. Music and the Psychology of Convention. Philadelphia: University of Pennsylvania Press.
Heinonen, Yrjö. 1992. Abstraktiset ja konkreettiset mallit The Beatles-yhtyeen musiikissa. [Abstract and Concrete Models in the Music of the Beatles.] In Etnomusikologian vuosikirja 4, 1991-92 (5-55).
Heinonen, Yrjö. 1995. Elämyksestä ideaksi — ideasta musiikiksi. Sävellysprosessin yleinen malli ja sen soveltaminen Beatles-yhtyeen laulunteko- ja äänitysprosessiin. [From Experience to an Idea — from an Idea to Music. A General Model of the Compositional Process and its Application to the Songwriting and Recording Process of the Beatles.] Contains an English summary. Jyväskylä: University of Jyväskylä.
Hermerén, G. 1975. Influence in Art and Literature. Princeton: Princeton University Press.
LaRue, J. 1961. Significant and Coincidental Resemblance between Classical Themes. Journal of the American Musicological Society (224-234).
Laske, O. 1989. Composition Theory: An Enrichment of Music Theory. Interface, vol. 18 (45-59).
Meyer, L. B. 1973. Explaining Music. Chicago: Chicago University Press.
Nelson, D. L. & Schreiber, T. A. & McEvoy, C. L. 1992. Processing Implicit and Explicit Representations. Psychological Review, vol. 99, No. 2 (322-348).
Platoff, John. 1988. Writing about Influences: Idomeneo, A Case Study. In Explorations in Music, the Arts, and Ideas. Essays in Honor of Leonard B. Meyer. Edited by Eugene Narmour and Ruth A. Solie (43-65). Stuyvesant: Pendragon Press.
Schank, R. C. & Abelson, R. 1977. Scripts, Plans, Goals and Understanding. Hillsdale (New Jersey): Lawrence Erlbaum Associates.
Tulving, E. 1983. Elements of Episodic Memory. New York: Oxford University Press.