Listener Preferences and Early Repetition in Musical Form
Joy Ollen and David Huron
School of Music, Ohio State University
ABSTRACT
The purpose of this study was to determine listeners’
preferences for patterns of successive repetition in musical
form. Twenty-nine musically sophisticated and naïve
participants performed a two-alternative forced-choice task in
which they selected the “most musical” composition from
pairs of stimuli that differed only with respect to form-related
repetition.
Compositions were comprised of two or three randomly-
selected discrete musical passages specifically composed for
the experiment. For each of the 44 trials, compositions of the
same length and using the same musical passages were paired
together. Paired compositions differed only in that one
exhibited patterns of successive repetition that decreased over
the course of the work and the other exhibited patterns that
increased. Results from a preliminary analysis indicate that
listeners prefer “early repetition” forms over “late repetition”
forms. These results are consistent with Huron and Ollen
(2004) who found evidence of early repetition in a cross-
cultural sample of 50 musical works.
1. BACKGROUND
Patterns of repetition in music are typically addressed in
discussions of musical form. The subject of form has attracted
extensive theoretical discussion and elucidation (Bairstow,
1943/1972; Berry, 1966; Davie, 1953/1966; Goetschius,
1904; Langer, 1953; Meyer, 1973; Riemann, 1910; Salzer,
1952; Schoenberg, 1937-1948/1967; de Stwolinski, 1977;
Tovey, 1944). Within the Western classical music tradition,
music theorists have identified many common forms. In other
cultures, discussions of musical form sometimes point to
similar notions of set structures (Fujie, 1996; Sachs, 1962;
Wade, 1979). Nicholas Cook has argued, however, that
stereotypes of various classical music forms were originally
formulated to teach composition and not to reflect how
listeners aurally experience a musical work (1987a). For
example, listeners appear to be less perceptive of large-scale
tonal closure as a work’s duration increases (Cook, 1987b) and
may judge the musical unity of a work quite differently based
upon their level of musical training (Tan, 1998).
In a correlational study, Huron and Ollen (2004) analyzed a
cross-cultural sample of 50 musical works for patterns of
repetition. The sampled works were drawn from five continents
and across five centuries. They were limited primarily to
instrumental works to avoid the potentially confounding
effects of text on musical form. Additionally, only works that
exhibited repetitive structures with identifiable units were
used; improvisatory and through-composed works were
omitted from consideration. The works were then analyzed by
assigning letter-names, beginning with A, to each unit of
repetition or “figure.” Thus, the analysis of A-A-B-A-A1-C
would correspond with a work whose opening figure was
played twice, followed by some contrasting material, a return
of the opening figure along with a varied repetition,
concluding with some previously unheard material. For these
analyses, figures did not represent the same structural level
within or between works; they merely denoted discrete musical
material that was either repeated elsewhere (the “A” material in
the previous example) or never recurred (the “B” and “C”
material). To put it in Western music theoretic terms, the
figures did not necessarily represent motives or phrases;
although in some cases, they might have.
Huron and Ollen tested two hypotheses and found significant
results for both. Of special interest to the present study was
their finding that longer sequences of repetition are more
likely to occur in the first half of a work than in the second
half. Using their analysis protocol, a musical work is more
likely to exhibit a structure like A-A-A-B-B-A-C-B-A than A-B-
C-B-B-B-A-A-A. Huron and Ollen dubbed this phenomenon
“early repetition.”
If musical works from many cultures and time periods exhibit
this early repetition, it is possible that, on some level, listeners
may expect musical works to be structured in this way.
Moreover, it would be interesting to know whether listeners
judge works exhibiting this pattern of decreasing repetition
over time as being “more musical” than works violating this
pattern.
1. AIMS
The purpose of the current experiment was to discover whether
listeners show a preference for compositions exhibiting early
repetition. In order to test the hypothesis, participants with
varying levels of formal musical training judged which
composition in each of 44 pairs sounded more “musical.”
Both compositions within each pair used the same musical
material, but differed in their patterns of successive
repetitions: one composition exhibited decreasing patterns of
successive repetitions (conforming to the findings of early
repetition in Huron & Ollen, 2004) and the other exhibited
increasing patterns (violating the findings). To anticipate the
results, participants showed a significant preference for the
compositions with decreasing patterns of successive
repetitions.
ISBN 1-876346-50-7 © 2004 ICMPC 405