Anda di halaman 1dari 8

Psychology of Aesthetics, Creativity, and the Arts

2007, Vol. 1, No. 2, 53 60

Copyright 2007 by the American Psychological Association


1931-3896/07/$12.00 DOI: 10.1037/1931-3896.1.2.53

Film Music: Are Award-Winning Scores and Songs Heard in Successful


Motion Pictures?
Dean Keith Simonton

This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

University of California, Davis


Using a sample of 401 feature-length narrative films released between 1998 and 2003, the current study
examines whether award-winning film music is more likely to appear in successful films. Film success
was measured using two measures of critical evaluations, a composite measure of best picture awards and
nominations, and box office gross, whereas the success of the film music was gauged by the number of
awards and award nominations received. In addition, control variables were defined for production costs,
release date, release season, runtime, MPAA rating, and genre (drama, comedy, romance, musical,
animation, and foreign-language). Although music awards and nominations were positively correlated
with film success, the score rather than song was primarily responsible for the relationship. Moreover,
after introducing the control variables, song awards had no relation whatsoever, whereas score awards
were still positively associated with the film success as measured by best-picture nominations and
awards.
Keywords: film music, scores, songs, awards, critics, box office

whole. Like the other components, the music was inserted, incorporated, or appended to enhance the films overall impact. For
example, because music is so effective in conveying and evoking
emotion (Juslin & Sloboda, 2001), both scores and songs are often
conceived to underline an affective mood, reaction, or condition,
whether joy or melancholy, triumph or defeat, tranquility or terror
(for review, see Cohen, 2001). Indeed, sometimes the director or
the producer has summarily rejected a composers music if it did
not serve the films cinematic goals, no matter how excellent the
music might be as a standalone creation. A notorious example is
Bernard Hermanns score for Alfred Hitchcocks Torn Curtain, its
rejection ending a director-composer collaboration involving nine
films, including the classics Vertigo, North by Northwest, and
Psycho. But what evidence is there that film music actually has a
beneficial consequence?
One possible item of evidence is that many organizations that
bestow annual honors for cinematic achievement include one or
more awards for film music. This applies to the Oscars for Best
Original Score and Best Original Song given by the Academy of
Motion Picture Arts and Sciences, the Golden Globes for Best
Original Score and Best Original Song awarded by the Hollywood
Foreign Press Association, and the Anthony Asquith Award for
Achievement in Film Music bestowed by the British Academy of
Film and TV Arts. Significantly, the Academy Awards for the two
film music categories began in 1935 (for films released in 1934),
just shortly after the advent of the sound erathe Oscar for Best
Sound was first given in 1931and actually before other awards
strongly associated with filmmaking (viz., editing in 1936, supporting male and female acting in 1937, film special effects in
1940, costume design in 1949, and makeup in 1968). Hence,
industry professionals have long recognized music to constitute a
critical feature of films.
Furthermore, award winning scores and songs include works created by an impressive diversity of eminent composers. Among the

Some of the worlds most played music was originally composed for film. Sometimes the music assumes the form of unforgettable songs, such as Over the Rainbow from the Wizard of
Oz, Moon River from Breakfast at Tiffanys, Mrs. Robinson
from The Graduate, On the Road Again from Honeysuckle Rose,
and My Heart Will Go On from Titanic. Other times instrumental music highlighted a significant character, scene, or moment,
such as The March of the Charioteers from Ben-Hur, the The
Murder in Psycho, Zorbas Dance from Zorba, the Greek, and
the Hymn to the Fallen in Saving Private Ryan. Yet other times
the music represents an orchestral composition that virtually epitomizes the entire motion picture. Examples include the themes
from Gone With the Wind, The Magnificent Seven, Exodus, Lawrence of Arabia, Doctor Zhivago, The Godfather, Jaws, Rocky,
Star Wars, Chariots of Fire, Dancing with Wolves, and Schindlers
List. So effective are these compositions that many are often heard
on concert performances and classical music radio stations alongside more traditional art music compositions by Bach, Mozart, and
Beethoven. Moreover, since 1959 outstanding film music has
received special recognition at the Grammy Award ceremonies,
joining not just classical but also rock, country, jazz, blues, rap,
and numerous other styles of musical expression.
As striking as this music may be, it must never be forgotten that
these pieces were once an integral part of a complex motion
picture. The music was just one component of numerous other
contributions, such as the screenplay, direction, acting, cinematography, art direction, costume design, special effects, and editing.
Its original purpose was always to be no more than part of the

Dean Keith Simonton, Department of Psychology, University of California, Davis, California.


Correspondence concerning this article should be addressed to Dean Keith
Simonton, Department of Psychology, One Shields Avenue, University of
California, Davis, CA 95616-8686. E-mail: dksimonton@ucdavis.edu
53

This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

54

SIMONTON

most distinguished are classical musicians (Aaron Copland, John


Corigliano, Tan Dun, Philip Glass, Erich Wolfgang Korngold, Toru
Takemitsu, John Williams, etc.), jazz, rock, and pop musicians (Phil
Collins, Neil Diamond, Bob Dylan, Duke Ellington, Eminem, Elton
John, Mick Jagger, Madonna, Barry Manilow, Paul McCartney,
Prince, Ryuichi Sakamoto, Carly Simon, Sting, Bruce Springsteen,
Vangelis, Stevie Wonder, etc.), musical theater composers (Jerome
Kern, Frederick Loewe, Alan Menken, Richard Rodgers, Stephen
Sondheim, Andrew Lloyd Webber, etc.), songwriters (Harold Arlen,
Burt Bacharach, Irving Berlin, Isaac Hayes, Giorgio Moroder, Randy
Newman, Cole Porter, Diane Warren, etc.), and, of course, film music
specialists both legendary (Elmer Bernstein, Bernard Hermann,
Maurice Jarre, Henri Mancini, Ennio Morricone, Alfred Newman,
Alex North, Miklos Rozsa, Max Steiner, Franz Waxman, Dimitri
Tiomkin, Victor Young, etc.) and contemporary (John Barry, Jerry
Goldsmith, Marvin Hamlisch, James Horner, Michel Legrand, Rachel
Portman, Zbigniew Preisner, Howard Shore, Hans Zimmer, etc.). As
if these names were not sufficient proof, each year since 1987 the
American Society of Composers, Authors and Publishers (ASCAP)
has been offering awards for the Most Performed Songs from Motion Pictures, The Most Performed Theme, and the film composers
of the Top Box Office Films.
In addition, an ample body of experimental research indicates how
music contributes directly to the cinematic experience. This music not
only helps establish the appropriate mood, emotion, or arousal (e.g.,
Ellis & Simons, 2005; Pfaus, Myronuk, & Jacobs, 1986), but also
facilitates the cognitive understanding of the films narrative course
(e.g., Boltz, 2001, 2004; Bullerjahn & Guldenring, 1994; Vitouch,
2001). Moreover, both music and film have comparable temporal
structures for the expression of dynamic transformations (e.g., a
beginning, middle, and an end, and the time-wise placement of
conflicts and their resolution). These parallels should enable music to
reinforce the progression of the visual and verbal content (for relevant
discussion, see Cohen, 2002; Rosar, 1994). Thus, there can be no
doubt that music perception influences visual perception (Iwamiya,
1994; Thompson, Russo, & Sinclair, 1994), a fact that can even be
demonstrated with respect to animated geometric figures (Marshall &
Cohen, 1988).
Notwithstanding this apparent empirical evidence, none of this
work directly addresses a fundamental question in cinematic aesthetics: is the musics standing as an independent aesthetic experience positively related to the films cinematic effectiveness or
impact? For example, although the experimental literature shows
that music has an impact on visual interpretation, it does not
indicate whether good music does a better job than mediocre
music. To be sure, one might argue that such a positive association
would be implicit in the empirical findings. But that is far from the
case. In fact, one could argue the opposite, namely that effective
film music might retreat to the background, providing subliminal
priming of cognitive and emotional associations. Music so outstanding that it becomes foreground could possibly divert attention
from the narrative, obliging audiences to pay attention to the music
itself rather than to its cinematic connotations. Great music could
be a distraction rather than a contribution.
What about all the movie awards bestowed by various organizations? It turns out that the evidence here is not as strong as first
meets the eye. For instance, the ASCAP Awards on closer scrutiny
are irrelevant to this issue. Honoring those composers whose films
are box office successes or those songs that are most frequently

performed does not indicate whether the composers or the songs


themselves had positive cinematic consequences. In particular, the
awards cannot rule out the possibility that the same composers
wrote even more exceptional music for box office failures or that
the popular songs appeared in films that critics considered
bombs or turkeys. Likewise, the bestowal of Oscars, Golden
Globes, and Asquith Awards cannot be taken as support unless it
is shown that these honors are most likely to be granted to movies
that also receive best picture awards. Yet the awards for best score
and best song are often given to films that have only one distinction, namely, that they have been nominated for having outstanding film music.
This disconfirming outcome has been demonstrated in several
recent studies. The first focused on the Oscars, looking at how well
the Academy Awards in the major categories predicted two separate criteria of cinematic success, the best picture award and the
number of stars the film received in movie guides (Simonton,
2002). Although both best score and best song awards exhibited
positive associations with best picture, these associations were
small and, in the case of songs, not always statistically significant.
In the instance of movie guide ratings, the score awards did not
make a significant contribution and, even worse, the song awards
were slightly negative predictors (see also Simonton, 2004a). In
another study, factor analysis determined how the awards given by
seven major organizations clustered together (Simonton, 2004b).
The score and song awards formed a separate musical cluster
that was largely orthogonal to the other factors, namely, the
dramatic (direction, screenplay, acting, and editing), visual (cinematography, art direction, costume, and makeup), and technical
clusters (visual effects, sound effects editing, and sound). Furthermore, although the musical cluster had a very small positive
association with the number of nominations and awards received
for best picture, the cluster had a negative correlation with the
ratings received in movie guides. Finally, another study showed
that although awards received for score and song are positively
associated with a films budgetit costs money to purchase the
services of acclaimed composers and songwriterssuch film music honors did not display significant correlations with best picture
awards, critical acclaim, movie guide ratings, or box office earnings, whether on the first weekend or the final gross (Simonton,
2005a). Taken together, this research suggests that award-winning
film music may not be associated with successful films. And, in
the case of songs, the association might even be negative.
This recent archival research is vulnerable to four major criticisms.
First, score and song awards were sometimes combined into a
single indicator even though the reliability coefficient for the
resulting composite was relatively low (Simonton, 2004b). Specifically, whereas the dramatic, visual, and technical clusters had
internal-consistency reliabilities of .88, .83, and .73, respectively,
the music clusters estimated reliability was only .55. Moreover,
this low reliability estimate may be partly ascribed to the fact that
the organizations used to define the composite measures did not
always include an award category for film music, an omission that
might suggest a certain bias against that cinematic component
(viz., the New York Film Critics Circle, the National Board of
Review, and the National Society of Film Critics).
Second, previous research used samples that excluded animated
films and films in non-English languages even though that exclu-

This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

FILM MUSIC

sion necessarily omitted many award-winning musical compositions (e.g., the animated films Lion King and Toy Story, and the
foreign-language films The Umbrellas of Cherbourg and Crouching Tiger, Hidden Dragon).
Third, no relevant study implemented the full set of control
variables suggested by previous research, such as production
budget, year and season of release, genre, timing, and Motion
Picture Association of America (MPAA) rating (Holbrook,
1999; Litman, 1983; Litman & Kohl, 1989; Simonton, 2005a,
2005b). As an example, it is certainly necessary to control for
genre given that prize-winning songs are highly likely to appear
in musicals, a genre that may not do well by many cinematic
standards (Simonton, 2005a).
Fourth and last, prior investigations failed to consider the fact
that the success of a film can be assessed by more than one
criterion and that these criteria do not necessarily correlate with
each other (Holbrook, 1999; Simonton, 2005b). Specifically, cinematic success can be evaluated four major ways: the evaluations
of critics at the time of the films release, the films box-office
earnings, the best-picture awards and nominations the film receives, and the evaluations that the film receives in movie guides
published a year or more after the completion of the films theatrical run.
Hence, the present investigation attempts to overcome the methodological limitations of the previous inquiries. Given more reliable assessments, more inclusive samples, more adequate controls,
and multiple success criteria, is it possible to conclude that successful movies are indeed more likely to feature award-winning
music?

55

such as direction, screenplay, and acting. Therefore, it may be


possible to assess the contribution of film music with the same
reliability as the other film components (because the reliability of
a composite is a positive monotonic function of the number of
items). Because the Online Film Critics Society did not start
offering awards for film music until 1999, the sample began with
films released in 1998. To ensure that the sampled films would
possess evaluations in movie guides, the sampling was terminated
at films released in 2003. Although animated and foreign-language
films were included, documentaries were excluded, thus confining
the sample to feature-length narrative films. It is rare for documentaries to feature outstanding musicthe song More from
Mono Cane constituting one of the rare exceptionsand in any
case the function of music in documentaries may be distinct from
that in a film that has a plot and characterization. After deleting the
five documentaries (none of which had received any recognition
for film music), the sample consisted of 401 films. This figure
amounts to about 67 films per year or more than a film per
weekand hence a high percentage of all widely released films in
a given year. However, because of missing values on certain
variables to be defined below, the analyses often were conducted
on a subset of this initial sample.
It should also be observed that the sampling period of 1998
2003 only partly overlaps those in previous investigations (1927
1999 in Simonton, 2002, 1968-1999 in Simonton, 2004b, 19752002 in Simonton, 2004a, 1997-2001 in Simonton, 2005a). Hence,
the investigation can be considered a replication and extension of
previous work. The earlier studies then provide a secure baseline
for evaluating the impact of the sampling and measurement improvements.

Method
With the exception of the movie-guide ratings, the raw data
came from electronic sources, especially such as the Internet
Movie Database (http://us.imdb.com/Sections/Awards/) and various official sites, such as the Academy Awards (http://
www.oscars.org/awards_db/ index.html). All correlations reported
throughout the remainder of this article are Pearson product
moment correlation coefficients (rs); all internal-consistency reliability coefficients are Cronbach alphas (s).

Sample
The sampling began by taking all films that were nominated for
awards in major film categories from the following organizations:
the Academy of Motion Picture Arts and Sciences (Oscars), the
Hollywood Foreign Press Association (Golden Globes), the
British Academy of Film and TV Arts (BAFTAs), the Los
Angeles Film Critics Association, the Chicago Film Critics
Association, the Broadcast Film Critics Association, and the Online Film Critics Society. That is, at least one of these organizations had to recognize the films achievement in one or more of the
following 17 categories: picture, direction, screenplay, lead male
and female actors, supporting male and female actors, film editing,
cinematography, art direction, costume design, makeup, visual
effects, sound effects editing, sound, and, of course, score and
song. These seven organizations were selected because they all had
at least one award category for film music, placing that category on
a par with the other major categories of cinematic achievement,

Measures
The measures fell into the following groups: movie awards, box
office gross, critic evaluations, and control variables.

Movie Awards
The nominations and awards given by the 7 organizations with
respect to the 17 cinematic categories were used to define a
corresponding set of 17 composite measures. For each composite,
2 points were assigned for an award, 1 point for a nomination, and
0 points for no recognition for that category for that organization.
Although this coding is clearly approximate, it provides a reasonable scaling according to past research on film samples comparable to the current sample. In the case of movie-guide ratings, the
number of stars added by an award over a mere nomination is
about as great as the number of stars added by a nomination over
no nomination whatsoever (Simonton, 2002, 2004a). A similar
pattern holds for box-office performance: the increment added by
an award over a nomination is about the same as that added by a
nomination over no nomination (Ginsburgh, 2003). Hence, assigning 2 points for an award is better than assigning any other integer
number of points.
In addition, points were added using the same 3-level scale
according to the nominations or awards received from 11 guild
associations germane to the same categories of cinematic accomplishment. These were the Producers Guild of America (picture),

This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

56

SIMONTON

Directors Guild of America (direction), the Writers Guild of America (screenplay), the Screen Actors Guild (all four acting categories), the American Cinema Editors (film editing), the American
Society of Cinematographers (cinematography), the Society of
Motion Picture and TV Art Directors (art direction), the Costume
Designers Guild (costume), the Motion Picture Sound Editors
(sound effects editing), and the Grammy Awards (score and song).
Previous research has shown that awards received from these more
specialized guild organizations are correlated with those bestowed
by the more general professional and critical societies on which the
sample was based (Simonton, 2004a, 2004b). For those categories
that were recognized by all 8 organizations, the composite score
could range from 0 (no awards or nominations whatsoever) to 16
(awards received from all organizations). For the less prominent
award categories, such as sound, fewer than 8 scores make up the
composite. The resulting composites had the following internalconsistency reliability coefficients (number of items in parentheses): picture .85 (8), direction .90 (8), screenplay .86 (8), lead male
acting .89 (8), lead female acting .89 (8), supporting male acting
.81 (8), supporting female acting .87 (8), film editing .79 (3),
cinematography .88 (7), art direction .79 (3), costume design .58
(3), makeup .78 (2), visual effects .72 (2), sound effects editing .40
(2), sound .75 (2), score .80 (8), and song .81 (4). These reliabilities are as good as or better than those found in previous studies
(Simonton, 2004a, 2004b). Especially notable for our current purposes are the highly respectable reliabilities for score and song,
which are higher than many other cinematic categories. This alone
justifies the change in the organizations used to define the sample
in comparison to previous investigations.
Because factor analytic results have shown that the awards tend
to form clusters (Simonton, 2004b), an additional set of composite
measures was generated from 16 of the above measures (keeping
the best picture assessment as a separate criterion variable). These
four creative clusters were (a) dramatic (direction, screenplay, lead
male and female actors, supporting male and female actors, and
film editing; .79; 7 items; M 0.63, SD 1.18); (b) visual
(cinematography, art direction, costume design, and makeup;
.82; 4 items; M 0.33, SD 0.85); (c) technical (visual effects,
sound effects editing, and sound; .79; 3 items; M 0.37,
SD 0.79); and (d) music (score and song; .65; 2 items; M
0.17, SD 0.39). Although the reliability of the music composite
is higher than found in previous studies (e.g., Simonton, 2004b), it
remains lower than the reliabilities of the separate score and song
composites. Also, although the two have a positive correlation, the
coefficient is just .49 ( p .001). This indicates that score and
song measures share less than 25% of their variance, a partial
independence that reflects the fact that the composer of the film
score is seldom identical to the composer of a film song (e.g.,
Howard Shore composed the score for Lord of the Rings: Fellowship of the Ring whereas the song May It Be was composed by
Enya and Nicky Ryan to lyrics by Roma Ryan). Hence, their
cinematic consequences should also be analyzed separately instead
of being united in a single measure of award-winning film music.
Below I analyze the data both ways to allow comparisons with past
research (e.g., Simonton, 2002, 2004b).
The foregoing awards measures are available for all 401 films in
the sample.

Box Office Gross


The U.S. gross earnings at the termination of a films run were
available for 387 cases. The figures were expressed to the nearest
million in U.S. dollars. The gross ranged from $431 million to
appreciably less than a half million (M 59.76, SD 77.33).
Because this variable exhibits considerable skew blockbusters
being extremely rare in comparison to the box office flopsit was
subjected to a logarithmic transformation. This transformation
renders the distribution more normal while also reducing the
impact of outliers (M 3.04, SD 1.75).

Critic Evaluations
Two distinct measures of critical acclaim were defined.
Metacritic evaluations. The website Metacritic (at http://
www.metacritic.com/) systematically compiles the judgments of
all major film critics, providing a weighted composite that ranges
along a 100-point scale. Because the site did not start the compilation of these scores until recently, the numbers are only available
for films released from the year 1999 on, or a total of 316 films.
The scores ranged over almost the full range of possible scores,
from a minimum of 14 to a maximum of 98 (M 68.01, SD
14.81). These scores are calculated shortly before, during, or
shortly after the films theatrical release, and hence they usually
represent critical judgments prior to the award ceremonies that
define the earlier measures. This priority does not apply to the
second assessment of critical acclaim.
Movie guide ratings. A composite index of film evaluations
was generated from five different movie guides (viz., Bleiler,
2004; Craddock, 2004; Maltin, 2004; Martin & Porter, 2004;
Walker, 2004). In these guides professional film critics provided
some rating system, most often using stars. Previous investigations have shown that these film evaluations exhibit a substantial
consensus (Boor, 1990, 1992; Holbrook, 1999; Simonton, 2004b).
Consequently, the ratings were converted into a quantitative measure by the following procedure. First, when necessary, each rating
was converted into a 5-star scale ranging from 5 (masterpiece or
classic) and 1 (turkey or bomb). If a film was not rated in a
particular guide, then it was assigned a missing value for the
corresponding score. The five assessments were then averaged
across all nonmissing values. The internal-consistency reliability
estimate for this measure was .79, which is about the same order
of magnitude as the award composites. Because these movie
guides contain ratings of released videos or DVDs, some films are
omitted (e.g., foreign-language films that are still doing the art
house circuit). As a consequence, missing values slightly reduced
the sample size to 393. The range in scores was substantial, namely
from 1- to 5-star films (M 3.45, SD 0.62).
It is noteworthy that the above two critic measures correlate .74
( p .001, n 313). Hence, the two assessments substantially
agree despite one being largely preaward and the other entirely
postaward in timing. This may be viewed as a kind of test-retest
reliability coefficient. As such it is notable that the coefficient is
about the same size as the internal-consistency reliability of the
movie-guide assessment. Subjective evaluations of films are
equally consistent and stable over time (see also Ginsburgh, 2003).

FILM MUSIC

This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

Control Variables
To remove possible artifacts that might undervalue the impact of
film music on cinematic success, several controls were defined.
These control variables were selected because previous investigations have established that they correlate with one or more criteria
of cinematic creativity (e.g., Simonton, 2004b).
Production budget. For 291 films it was possible to obtain the
production costs expressed in units of a million U.S. dollars,
rounded off to the nearest million. The actual range was from less
than half of a million to $175 million (M 47.90, SD 38.77).
This measure was also subjected to a logarithmic transformation to
accommodate the extreme skew in the distribution (M 2.36,
SD 1.26). It should be recognized that production costs are
considered proprietary information and therefore they are most
often supplied as rounded estimates rather than as exact figures.
Release date. This variable was simply the year in which the
film was first widely distributed in U.S. movie theaters. Because
the films only span an 8-year period, this variable simply adjusts
for extremely short-term trends, such as a momentary rise in
production costs or a temporary decline in movie guide ratings.
Release season. Theatrical releases tend to be carefully timed to
maximize potential profits and award eligibility (Litman, 1983;
Litman & Kohl, 1989). As a consequence, three variables were
defined to indicate the season in which each film first appeared. In
particular, zero-one dummy variables were defined for three standard
release periods: Easter (March and April; 13%), Summer (May
through August; 29%), and Christmas (November and December;
31%). Any film released at some other time of the year defined the
intercept (or comparison group) in the multiple regression equation.
Runtime. This variable was defined by the number of minutes
that it usually takes to show the film. If the various sources
disagreed regarding the duration (e.g., original release vs. directors cut), then the median timing was used. This information was
available for 393 films (M 118.29, SD 47.76).
MPAA ratings. A final set of zero-one dummy variables recognized the audience ratings published by the Motion Picture Association of America. In particular, the dummy variables coded whether or
not the film was rated G (4%), PG (10%), PG-13 (30%), and R (50%).
These four categories account for 94% of the sample. It should be
noted that many non-English language films are not rated.
Film genre. Another set of zero-one dummy variables were
defined to record the most common genre according to the categorizations provided by the Internet Movie Database, the most
comprehensive categorization available in any form, print or electronic (Simonton, 2004b). These dummy variables specifically
coded whether the film was a drama (68%), comedy (34%),
romance (20%), or musical (3%). These four categories are not
mutually exclusive given that a particular film might belong to
more than one genre (e.g., musical comedies, romantic dramas).
Unfortunately, more refined genre definitions were not possible
(e.g., mysteries, westerns, science fiction, fantasy) because the
more specialized genres accounted for only a small percentage of
the films in the sample and usually overlapped considerably with
other categories (e.g., crime and film-noir, western and drama),
thus severely reducing their utility as genre identifiers. Nevertheless, because of the nature of the current sample, two additional
zero-one dummy variables were used to define films that fit the
following classifications: animation (7%) and foreign language

57

(20%). The latter was assigned to any film that was not in English
(whether British, American, Australian, etc.).

Results
Table 1 shows the correlations between the creative clusters of
dramatic, visual, technical, and music and the criteria of cinematic
success, namely, the two critic evaluations (metacritic and movie
guide), the best picture awards composite, and box office gross. In
addition, the music cluster has been split into its separate components,
score and song, to produce another set of correlations. The following
conclusions should be apparent at once. First, the music cluster
correlates with all four criteria; the highest correlation is with best
picture awards and the lowest with the metacritic ratings, the movie
guide ratings and gross earnings coming between. Second, the impact
of film music is similar to the other creative clusters, albeit the
dramatic cluster is the more consistently strong correlate (see also
Simonton, 2005b). Third, of the two components of the music cluster,
it is the score that is most strongly associated with the criteria. With
the exception of gross, where score and song have about the same
influence, score has about twice the impact as song, and in the case of
the metacritic scores song lacks a statistically significant relationship
( p .08). It is for this reason that the correlations between score and
the success criteria tend to be higher than those between the music
cluster and the same criteria, again with the sole exception of box
office performance. Hence, it seems safe to infer that by most success
criteria award-winning film music tends to be positively linked with
successful films, this linkage being especially strong in the case of the
score.
Nonetheless, the above conclusion must be tempered by the realization that these correlations do not incorporate any statistical controls. Indeed, the associations do not even correct for the influence of
the other award clusters. Therefore, multiple regression analysis was
used to implement the desired correction. In particular, each of the
four success criteria was regressed on the four creative clusters plus
the control variables for budget, release date, release season, timing,
MPAA rating, and genre (including animation and foreign language).
Furthermore, the four regressions were repeated with the music cluster split into its separate score and song components. The standardized
partial regression coefficients (betas) from these analyses are reported in Table 2.

Table 1
Pearson Product-Moment Coefficients Between Cinematic
Success Criteria and the Four Creative Clusters With and
Without the Music Cluster Separated Into Score and Song
Critic evaluations
Composite variable
Metacritic Movie guide
Creative cluster
Dramatic
.38***
.50***
**
Visual
.15
.27***
Technical
.04
.17***
Music
.18**
.26***
Separate music categories
Score
.22***
.31***
Song
.10
.15**
n
316
393
*

p .05.

**

p .01.

***

p .001.

Best picture Gross

.67***
.51***
.38***
.46***

.24***
.29***
.43***
.34***

.56***
.25***
401

.28***
.30***
387

SIMONTON

58

Table 2
Standardized Partial Regression Coefficients for Cinematic
Success Criteria as a Function of the Four Creative Clusters
With and Without the Music Cluster Separated Into Score and
Song

This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

Critic evaluations
Composite variable
Metacritic Movie guide
Creative cluster
Dramatic
.38***
.49***
Visual
.00
.00
Technical
.13
.14*
Music
.15*
.09
Separate music categories
Score
.14
.07
Song
.06
.04
R2
.46***
.48***
n
232
291

Best picture Gross

.63***
.15***
.12**
.13***

.19***
.01
.08
.08

.18***
.01
.79***
291

.01
.07
.65***
289

Note. The estimated partial regression coefficients include the following


control variables: production costs, release date, release season (Christmas,
summer, and Easter), runtime, MPAA rating (G, PG, PG-13, and R), genre
(drama, comedy, romance, and musical, plus animation and foreign language). Because the sample size is very large relative to the number of
independent variables, the adjusted-R2 for each equation is virtually identical to the R2 given in the table.
*
p .05. ** p .01. *** p .001.

It should be evident that the introduction of controls radically


changes the results. Now the only creative cluster that displays consistently strong correlations with all four criteria is the dramatic (see
also Simonton, 2004b, 2005b). By comparison, the music cluster
correlates only with metacritic assessments and best picture awards.
Furthermore, the outcome for the music cluster actually deteriorates
when it is split into score and song; only score has a positive correlation, and then solely with a single criterion, namely best picture
awards (albeit the coefficient for metacritic has a p .08). In contrast,
songs association with best picture honors is practically zero. Thus,
the impact of film music seems even smaller once necessary controls
are introduced into the regression equation. Even so, in no analysis is
the association negative, unlike what has been found in some prior
inquiries (Simonton, 2002, 2004a, 2004b; cf. Simonton, 2005a).
This more positive outcome may partly reflect the manner in which
the films were sampled. Besides basing the sample on the films
honored by seven organizations that specifically recognized music,
the sample included both animations and foreign-language films that
can often be sources of award-winning scores and songs. Accordingly, it is instructive to examine what happens when the latter two
types of films are removed from the sample. Table 3 provides the
results. The consequence differs very slightly from what is seen in
Table 2. Although the music cluster no longer correlates significantly
with the metacritic scores, its association with best picture awards
actually increases slightly, and not just for the music cluster but also
for the score component. Once more song proves to be irrelevant by
any of the more obvious standards used in this investigation. Still, the
correlations are again simply zero rather than significantly negative.
This implies that the previous studies that found negative associations
may have had an antisong sampling bias introduced by including
organizations that did not even recognize film music as a category of
cinematic achievement.
By the way, when all musicals are also deleted in the sample, the
outcome is virtually identical to what is shown in Table 3. This

result is probably not surprising given that this genre accounts for
only about 3% of the sample.

Discussion
This investigation obtained some justification for concluding
that award-winning music and successful films are inclined to go
together. Nominations and awards for film scores and songs are
positively correlated with two distinct critical evaluations, best
picture honors and box office earnings. However, half of these
associations vanish when variables are inserted in the regression
equations to control for various artifacts, and only one association
remains if the sample of animations and foreign-language films are
removed from the sample. Hence, the only assertion that can be
expressed with absolute confidence is that award-winning film
music, and especially highly honored scores, is positively related
to best picture awards. Moreover, although this relationship is not
as strong as the dramatic cluster (direction, screenplay, acting, and
editing), it is about the same magnitude as the visual cluster
(cinematography, art direction, costumes, and makeup) and the
technical cluster (visual effects, sound effects editing, and sound).
In this respect, the film composer could be considered as important
as most of the others contributing to the filmmaking process. As a
consequence, when consumers identify film music as one of the
criteria they use for deciding what films to see (Linton & Petrovich,
1988) they are not applying irrational standards. This criterion will
not systematically guide them toward an inferior cinematic experience as judged by critical evaluations and box-office success.
Furthermore, the criterion will make them more likely to see a
noteworthy picture as assessed by one reasonable standard how
many best picture awards and nominations it receives.
Nonetheless, two complications about this contribution deserve
discussion. The first has to do with why it is the score rather than the
Table 3
Standardized Partial Regression Coefficients for Cinematic
Success Criteria as a Function of the Four Creative Clusters
With and Without the Music Cluster Separated: Animations and
Foreign- Language Films Deleted
Critic evaluations
Composite variable
Metacritic Movie guide
Creative cluster
Dramatic
.42***
.53***
Visual
.02
.07
Technical
.13
.17**
Music
.10
.02
Separate music categories
Score
.13
.04
Song
.01
.01
R2
.46***
.53***
N
203
252

Best picture Gross

.69***
.11**
.11**
.18***

.23***
.02
.09
.01

.20***
.04
.82***
252

.03
.03
.63***
252

Note. The estimated partial regression coefficients include the following


control variables: production costs, release date, release season (Christmas,
summer, and Easter), runtime, MPAA rating (G, PG, PG-13, and R), genre
(drama, comedy, romance, and musical, plus not animation and foreign
language). Because the sample size is very large relative to the number of
independent variables, the adjusted-R2 for each equation is virtually identical to the R2 given in the table. Also, the R2 for Best Picture separate
categories is .83 ( p .001).
*
p .05. ** p .01. *** p .001.

This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

FILM MUSIC

song that features this asset. It should be clear that this cannot be
ascribed to the effects of genre, such as musicals, because the irrelevance of film songs is even more pronounced after introducing the
necessary controls. So the difference must reside with the distinctive
place that scores and songs have with respect to the overall cinematic
effect. It is possible that the scores, or soundtracks, are more likely to
enhance the narrative flow of the film, whereas songs may be just as
likely to distract as to contribute, so that the general influence across
a wide range of films is neutral. The distraction may be likely when
the song seems to have no reason for being there except to allow a
stara popular singer turned actor perhapsto show off his or her
latest hopeful hit. Meanwhile, the film itself momentarily grinds to a
halt, the cognitive and emotional direction lost, and the viewer forced
to stumble back on track after the disruption. Future experimental
research may help resolve this question by directly comparing the
effectiveness of instrumental and vocal music.
The other complication is that the positive impact of film music is
only consistently evident for a single criterion, namely best picture
awards and nominations. To account for this complication, let us first
deal with two methodological interpretations. One possibility is that
the variance is extremely truncated in all film success criteria except
best picture awards. However, this explanation appears very implausible. The two critical evaluations exhibited the full range or nearly so
(1 to 5 stars for movie-guide ratings and 14 98 points for metacritic
ratings) and the variation in the box-office criterion was no less so
(from less than $500,000 to nearly $500,000,000). Thus, the variance
in the criterion variables was minimally reduced. Another methodological interpretation is that the discrepant correlations merely represent method effects. That is, the same professionals, journalists,
and critics involved in the best picture nominations and awards are
often those engaged in the corresponding best score and song honors.
However, because many critics who belong to the Los Angeles Film
Critics Association, the Chicago Film Critics Association, the Broadcast Film Critics Association, and the Online Film Critics Society
produce reviews that are compiled in the metacritic measure, it is hard
to see why the latter does not also share method variance with the film
music assessment. Hence, both of these more mundane interpretations
can probably be dismissed.
An alternative and more provocative interpretation requires that we
recognize two facts. First, box office success has no significant
correlation with movie guide ratings (r .03, p .5) and best picture
awards (r .09, p .1) and a slightly negative correlation with
metracritic ratings (r .16, p .01). These correlations fall in line
with earlier studies that have established the distinction, even opposition, between film as art and film as business (Holbrook, 1999;
Simonton, 2005a, 2005b). The two types of films are highly distinguishable regarding genre, screenplay, the role of special effects and
big stars, and a host of other cinematic attributes. So it should come
as no surprise that award-winning film music bears no connection
with a motion pictures financial success.
The second fact concerns the correlation between best picture
awards and the two critic evaluations: these are .44 ( p .001) for
metacritic and .50 ( p .001) for movie guidessubstantially lower
than the .74 noted earlier. In more concrete terms, the critical assessments share less than half as much variance with best picture awards
as they do with each other. For reasons already noted, it cannot be said
that the critics differ from those in the organizations who provided
best picture awards and nominations; critics organizations were included in the composite measure. Furthermore, the high reliability

59

coefficient for this 8-item composite suggests the existence of a strong


consensus. Therefore, some other contaminant undermines the degree
of correspondence. One possibility is that the critic evaluations and
the best picture awards focus on different portions of the distribution
in film impact and assess contrasts in film impact in distinct ways.
Metacritic and movie guide ratings are engaged in discriminating the
full range of cinematic merit, from the turkeys or bombs to the
genuine masterworks of cinema. Best picture awards, in contrast,
distinguish the best of the best. Those films that are nominated for best
picture are highly likely to receive nominations in many other film
categories. Moreover, because the sample is confined to films that
have been nominated for at least one award in a standard category, the
best picture awards are better placed to make fine distinctions than are
the critic evaluations. In light of these differences, it is conceivable
that the film score is mostly responsible for helping convert a good
film to a great filmputting an already accomplished motion picture
well ahead of the pack. If so, then award-winning film music would
have more influence on best picture awards and nominations than on
critical evaluations.
Naturally, more empirical research is necessary before these issues
can be fully resolved. In addition, it would be important to explore
other aspects of the phenomenon ignored here. For instance, this
investigation has concentrated exclusively on music composed specifically for a particular film. But ever since the late 1960s, some
soundtracks began increasingly to consist of compilations from other
preexisting sources. A classic example is the music for the 1968 film
2001: A Space Odyssey. The director Stanley Kubrick originally had
asked Alex North, his composer for Sparticus, to write the score. But
while shooting the film, the director decided to play classical music on
the set to put everyone in the desired mood. Kubrick was so pleased
with the result that the final soundtrack consisted solely of compositions by Johann Strauss, Richard Strauss, Aram Khachaturyan, and
Gyorgy Ligeti, leaving North to publish his creative efforts as film
music without a film! A year later, in 1969, Easy Rider featured an
entire soundtrack composed of popular songs by, among others,
Steppenwolf, The Byrds, Jimi Hendrix, and Bob Dylan, thereby
establishing the same approach with respect to popular music. In fact,
the only song Dennis Hopper specifically commissioned for the
filmin this case by Stephen Stillswas not even used, and, like
Norths music, ended up being recorded separately on a Crosby, Stills,
Nash, and Young single. Whatever the specific examples, does
coopted film music have the same status as music tailored for a
specific film? Or is it less effective, its peripheral origins and its very
familiarity interfering with comprehension of a film as a new cinematic experience?
One unique feature of these compilation soundtracks is that the
music can be evaluated for aesthetic impact independent of the films
in which they appear. For instance, classical music can be assessed
according to the performance and recording frequencies (see, e.g.,
Simonton, 1980, 1995, 1998). Given these independent assessments,
the questions addressed in the current study can be reiterated. Will the
differential effectiveness of the adapted music correlate positively
with the impact of the film? Will the correlation be more conspicuous
for instrumental music rather than for vocal music? And, finally, will
the correlation also be restricted to best picture awards? Or will the
impact extend to critic evaluations and even box office earnings?
These questions can be answered using archival methods like those
used here. It would be equally valuable to conduct experimental
research that directly compares the impact of commissioned film

SIMONTON

60

music with music appropriated from noncinematic sources. For example, it would be instructive to compare directly Norths music for
2001: A Space Odyssey with that pieced together by Kubrik. The
comparison can operate both for the picture as a whole and on the
level of individual scenes (e.g., Norths Fanfare vs. the famous one
excerpted from the opening to Richard Strausss Also Sprach
Zarathustra; see Lipscomb & Kendall, 1994, for a possible experimental paradigm for addressing this question). The upshot of both
correlational and experimental studies should be a much richer appreciation of film music than we presently possess.

This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

References
Academy Awards Database. (n.d.). Retrieved from http://www.oscars.org/
awards_db/ index.html
Bleiler, D. (Ed.). (2004). TLA film and video guide: The discerning film
lovers guide 2005. New York: St. Martins Press/Griffin.
Boltz, M. G. (2001). Musical soundtracks as a schematic influence on the
cognitive processing of filmed events. Music Perception, 18, 427 454.
Boltz, M. G. (2004). The cognitive processing of film and musical soundtracks. Memory & Cognition, 32, 1194 1205.
Boor, M. (1990). Reliability of ratings of movies by professional movie
critics. Psychological Reports, 67, 243257.
Boor, M. (1992). Relationships among ratings of motion pictures by
viewers and six professional movie critics. Psychological Reports, 70,
10111021.
Bullerjahn, C., & Guldenring, M. (1994). An empirical investigation of
effects of film music using qualitative content analysis. Psychomusicology, 13, 99 118.
Cohen, A. J. (2001). Music as a source of emotion in film. In P. Juslin &
J. Sloboda (Eds.), Music and emotion: Theory and research (pp. 249
272). New York: Oxford University Press.
Cohen, A. J. (2002). Music cognition and the cognitive psychology of film
structure. Canadian Psychology, 43, 215232.
Craddock, J. (Ed.). (2004). VideoHounds golden movie retriever 2004.
Boston: Visible Ink.
Ellis, R. J., & Simons, R. F. (2005). The impact of music on subjective and
physiological indices of emotion while viewing films. Psychomusicology, 19, 15 40.
Ginsburgh, V. (2003). Awards, success and aesthetic quality in the arts.
Journal of Economic Perspectives, 17, 99 111.
Holbrook, M. B. (1999). Popular appeal versus expert judgments of motion
pictures. Journal of Consumer Research, 26, 144 155.
Internet Movie Database (n.d.). Accessed from http://us.imdb.com/
Sections/Awards/
Iwamiya, S. (1994). Interactions between auditory and visual processing
when listening to music in an audio-visual context: I. Matching II. Audio
quality. Psychomusicology, 13, 133154.
Juslin, P., & Sloboda, J. (Eds.). (2001). Music and emotion: Theory and
research. New York: Oxford University Press.
Linton, J. M., & Petrovich, J. A. (1988). The application of the consumer
information acquisition approach to movie selection: An exploratory
study. Current Research in Film, 4, 24 45.

Lipscomb, S. D., & Kendall, R. A. (1994). Perceptual judgement of the


relationship between musical and visual components in film. Psychomusicology, 13, 60 98.
Litman, B. R. (1983). Predicting success of theatrical movies: An empirical
study. Journal of Popular Culture, 16, 159 175.
Litman, B. R., & Kohl, L. S. (1989). Predicting financial success of motion
pictures: The 80s experience. Journal of Media Economics, 2, 3550.
Maltin, L. (Ed.) (2004). Leonard Maltins 2005 movie and video guide.
New York: Signet.
Marshall, S. K., & Cohen, A. J. (1988). Effects of musical soundtracks on
attitudes toward animated geometric figures. Music Perception, 6, 95
112.
Martin, M., & Porter, M. (2004). Video movie guide 2005. New York:
Ballantine Books.
Metacritic (n.d.). Accessed from http://www.metacritic.com/
Pfaus, J. G., Myronuk, L. D., & Jacobs, W. J. (1986). Soundtrack contents
and depicted sexual violence. Archives of Sexual Behavior, 15, 231237.
Rosar, W. H. (1994). Film music and Heinz Werners theory of physiognomic perception. Psychomusicology, 13, 154 165.
Simonton, D. K. (1980). Thematic fame, melodic originality, and musical
zeitgeist: A biographical and transhistorical content analysis. Journal of
Personality and Social Psychology, 38, 972983.
Simonton, D. K. (1995). Drawing inferences from symphonic programs:
Musical attributes versus listener attributions. Music Perception, 12,
307322.
Simonton, D. K. (1998). Fickle fashion versus immortal fame: Transhistorical assessments of creative products in the opera house. Journal of
Personality and Social Psychology, 75, 198 210.
Simonton, D. K. (2002). Collaborative aesthetics in the feature film:
Cinematic components predicting the differential impact of 2,323 Oscarnominated movies. Empirical Studies of the Arts, 20, 115125.
Simonton, D. K. (2004a). Film awards as indicators of cinematic creativity
and achievement: A quantitative comparison of the Oscars and six
alternatives. Creativity Research Journal, 16, 163172.
Simonton, D. K. (2004b). Group artistic creativity: Creative clusters and
cinematic success in 1,327 feature films. Journal of Applied Social
Psychology, 34, 1494 1520.
Simonton, D. K. (2005a). Cinematic creativity and production budgets:
Does money make the movie? Journal of Creative Behavior, 39, 115.
Simonton, D. K. (2005b). Film as art versus film as business: Differential
correlates of screenplay characteristics. Empirical Studies of the Arts, 23,
93117.
Thompson, W. F., Russo, F. A., & Sinclair, D. (1994). Effects of underscoring on the perception of closure in filmed events. Psychomusicology,
13, 9 27.
Vitouch, O. (2001). When your ear sets the stage: Musical context effects
in film perception. Psychology of Music, 29, 70 83.
Walker, J. (Ed.). (2004). Halliwells film and video guide 2005. New York:
HarperCollins.

Received May 19, 2006


Revision received May 19, 2006
Accepted May 30, 2006

Anda mungkin juga menyukai