Silent Strategies: Audiovisual Functions of the Music for Silent
Cinema
Marco Bellano (Università degli Studi di Padova)
Abstract
Studies on film music have often overlooked the difference between the
audiovisual strategies of sound cinema and the ones of silent cinema. However,
there are at least two audiovisual strategies which are peculiar to silent cinema: a
›bridge‹ function born from the improvisational nature of silent film music
practice, and a ›interdiegetic‹ function, which takes advantage of the impossibility
to hear the sounds of the world seemingly positioned beyond the silver screen. This
paper comments upon these two strategies. A succinct review of the literature that
already acknowledged the existence of these strategies, mostly in an indirect way
(from Ricciotto Canudo to Sebastiano Arturo Luciani, Edith Lang and George
West) leads to the discussion of examples from historical musical illustrations of
silents (e.g. one by Hugo Riesenfeld for Cecil B. DeMille’s CARMEN, USA 1915)
as well as from contemporary ones (e. g. Neil Brand’s 2004 music for THE CAT
AND THE
CANARY, USA 1927, Paul Leni). The change in reception conditions of
silent films between the early 20 th century and the present days is certainly
relevant; however, this paper does not aim to offer an insight into cultural-historical
context and reception, but to point out how the silent film language invited
composers in different periods to develop a set of audiovisual strategies that are
identical on a theoretical level.
Kieler Beiträge zur Filmmusikforschung, 9, 2013 // 46
The renaissance of silent film studies that started more than thirty years ago 1
had only a minor impact on film music theory. Silent film music has often
been regarded just as a vehicle of a ›primitive‹ audiovisual aesthetic,
waiting for the ›evolution‹ into sound film music (Lissa 1965, 98). As Tom
Gunning wrote this is »a biological and teleological logic« (Gunning 1996,
71) that conceives the later styles of cinema as »a sort of natural norm that
early cinema envisioned but was not yet capable of realizing because of
technological and economic immaturity and a natural need for a period of
development guided by a method of trial and error« (Gunning, 1996, 71-72).
However, whereas contemporary studies about the visual aspects of cinema
consistently dismiss such a logic, following the work of theorists such as
Gunning, André Gaudreault, Richard Abel, Noël Burch and Charles Musser,
the field of film music studies is still indulging in this old perspective. At
least, Rick Altman said in 2004 that »[i]t is time to include sound in silent
cinema’s historiographical revival« (Altman 2004, 9); however, no one
spoke about such a revival in the studies about theory and aesthetics of
silent film music.
1
The main events that marked the beginning of this renaissance were the 1978 FIAF
international conference Cinema 1900-1906 and the presentation of Kevin
Brownlow’s reconstruction of Abel Gance’s NAPOLÉON (France, 1927), in two
versions. The first one, produced by the British Film Institute and by the Images
Film Archives in association with Thames Television, had a projection speed of 20
frames per second (fps) and a new original score by Carl Davis; it had its debut at
the Telluride Film Festival in September 1979, and it was then reprised on
November 30, 1980 at the London Film Festival. The second version had a
projection speed of 24 fps and music by Carmine Coppola; it debuted at the Radio
City Music Hall in New York in January 1981. See Brownlow 1981;
Holman/Gaudreault 1982; Carl Davis in Brand 1998, p. 93.
Kieler Beiträge zur Filmmusikforschung, 9, 2013 // 47
1. Historical descriptions of audiovisual strategies
specific to silent films
The present instruments of film music theory surely allow to delve deeper
into the field of the music for silent films. It is true that many audiovisual
strategies are shared between silent and sound cinema. However, the
different nature of the two kinds of audiovisual entertainment gives
profoundly different meanings to strategies which seem to be identical only
on a superficial level.
Two peculiar audiovisual strategies will be discussed here: a ›bridge‹
function, born from the improvisational roots of silent film music practice
(which in its maturity relied more on compilation, though, while
improvisation itself was based on a sort of extemporary ›compilation‹ of a
known musical repertoire), and an ›interdiegetic‹ function, which takes
advantage of the impossibility to hear the sounds of the world seemingly
positioned beyond the silver screen. Both these strategies have a common
purpose: to create unity in the discourse of the images. They seem to
respond to a necessity to counterbalance the ›fragmentary‹ aspect of the
spectacle which, even before the advent of the classic style of montage, was
a primary concern for filmmakers and audiences. Cinema was fragmentary
even in the age of the ›single-shot‹ films (approximately 1895-1903),
because it selected a portion of the visual space. The various and notorious
accounts of peasants scared to death by the vision of heads floating in the
dark of the cinema hall could be remembered as funny yet meaningful
evidence of this problem.
My exploration starts from a succinct review of the historical literature that
already acknowledged the existence of these strategies, mostly in an indirect
way. Examples of these strategies will be given by quoting historical
Kieler Beiträge zur Filmmusikforschung, 9, 2013 // 48
musical illustrations of silents as well as contemporary ones, which,
notwithstanding the relevant differences in the way silent cinema is today
received and understood by audiences, still invite composers to develop
audiovisual devices which seem identical to the ones used by historical
composers, on a theoretical level.
It is true that, especially during the early period (1895 – ca. 1907), film
music did not spoil the audience with refined examples of audiovisual
interactions. Even if the popular story of the birth of film music 2 − because
of the necessity to hide the noise coming from the Geneva drive mechanism
and the sprockets of the projector − is today regarded just as a legend
(Simeon 1995, 18-19), many accounts report how often music stayed in the
cinema hall just to entertain the ear, without paying much attention to the
moving images. In a 1913 editorial of the monthly magazine The
Metronome, a cinema pianist from London admitted that some of his
colleagues would »simply strum a waltz or rag-time through, and go on to
the end of it, whether people in the pictures are dying or marrying […]«
(Anderson 2004, 175).3 Actually, many of the venues where cinema was
screened during the silent age could not afford high-level professionals at
the piano or holding the baton (Simeon, 1995, 117), even during the
›maturity‹ of the silents, when big cities in Europe and, especially, in
America, hosted deluxe musical presentations in extravagantly baroque
movie palaces (Beynon 1921, 13; Hampton 1970 [1931], 172; Herzog 1981,
15; Anderson 1988, xv-xxvi). While these expensive presentations often
relied upon original compositions or lush and eccentric compilations of
2
The work that is at the origin of the diffusion of this legend is London, 1970 [1936],
28.
3
Development of the Picture-Pianist. In: The Metronome 29:7 (July 1913).
Kieler Beiträge zur Filmmusikforschung, 9, 2013 // 49
staples from the symphonic repertory (Cavalcanti 1939, 100; Anderson
1988, xxiv-xxv), as was the case at the Regent, Strand and Capitol theatres
in New York, the routine in the smaller centers consisted of compilations of
piano pieces (on the basis of repertoires or cue sheets), or of free
improvisation, which did not mean free extemporary creation of music but,
more likely an extemporary assemblage of pieces from a certain musician’s
repertoire. When the word ›improvisation‹ is used to speak about the silent
film music practice, it should always be understood with the meaning
explained by Sergio Miceli (as it is in this article):
Everyone relied upon their own repertoire, reading the
music or playing by heart, with little and more or less
questionable adjustments to the film. So it should not be
excluded that the definition [of improvisation], especially
when used by a musician, was used as a reference more
to the etymological root than to the musical practice,
identifying
thus
improvisational
a
performance
mindset,
that
is
staged
to
say
with
an
without
preliminary preparation. (Miceli, 2009, 40)
Witnessing the prevalence of this kind of ›improvisational‹ practices, which
did not possess the ability to relate to the images in a precise way, several
early theorists were induced to underline how film music did not really have
any further function beyond the creation of a distraction from the otherwise
unreal silence of the spectacle. In 1913, Ernst Bloch argued that the blackand-white world shown by cinema had the »lugubrious appearance of a
Kieler Beiträge zur Filmmusikforschung, 9, 2013 // 50
solar eclipse«4 (quoted after Simeon 1995, 19): it was a mute reality,
deprived of many sensorial perceptions. The music, in his opinion, was there
just to »provide a substitution of all the missing senses« (ibid.). Later, Béla
Balázs indirectly agreed with Bloch, when he said that
a great part of the audience, in the cinema hall, is not
aware of the music; its existence is acknowledged just
when it stops. From a psychological point of view, this
phenomenon can be explained as follows: the human
being does not normally perceive the reality with a single
sense. The things we only see, we only hear, etc., do not
have the aspect of a three-dimensional reality (Balázs,
1975 [1948], 328; my translation).
So, for the sake of these basic purposes a continuous improvisation, even if
not conceived carefully to match or illustrate the images, would have been
enough. However, improvisation in the literal sense might not be the only
factor responsible for comments such as the ones by Bloch and Balázs. In
fact, the influence of improvisational practices also on the other musical
routines of silent cinema was a relevant one: so, it was possible to perceive a
certain ›detachment‹ between music and images also in, for example,
written compilations. Even the original scores that started to be composed
since 1905-19085 and flourished during the 1920s retained some traits of
4
Bloch, Ernst (1913) Über die Melodie im Kino. In: Die Argonauten (my translation).
5
Among the first films to be paired with original scores were LA MALIA DELL’ORO
(Italy 1905, Gaston Velle), NOZZE TRAGICHE (Italy 1906, Gaston Velle) and
ROMANZO DI UN PIERROT (Italy 1906, Mario Caserini, Filoteo Alberini, Dante
Santoni), with music by Romolo Bacchini, that preceded Camille Saint-Saëns’ work
Kieler Beiträge zur Filmmusikforschung, 9, 2013 // 51
improvisation. The fact that silent film music was prevalently a matter of
live performances reinforced the connection with improvisation. As Philip
Alperson argued,
musical improvisation can be appreciated for some of the
same values as can the action of musical performances in
the conventional musical situation. In particular, one can
appreciate the improviser’s sensitivity, lyricism and
general virtuosity as an instrumentalist or vocalist which
we associate with the narrower sense of musical
performance (Alperson 1984, 23).
This means that a performance always offers the musician a certain degree
of freedom, freedom that could be used to add something to the information
contained in the written score which is being played. However, this is just
the weaker bond between written silent film music and the practice of
improvisation. In fact, to improvise for the silents, as mentioned before,
usually involved a creative assemblage of pieces, generated from the
memory and sensibility of the player while the film was screened. The use
of fragments from a repertory is absolutely not in contrast with the idea of
improvisation: as Alperson noticed, »learning to improvise is often, in large
part, learning to master that tradition. Jazz musicians, for example,
frequently begin to learn to improvise by listening to and copying […] other
players’ musical phrases […] many of which have long ago attained the
status of formulae […]« (Alperson 1984, 22). The key feature of this basic
for L’ASSASSINAT DU DUC
Bargy). See Redi 1999, 54.
DE
GUISE (France 1908, André Calmettes, Charles Le
Kieler Beiträge zur Filmmusikforschung, 9, 2013 // 52
aspect of the improvisation is thus repetition: of a phrase, of a melody or of
a longer formula within a piece, or in different pieces referring to similar
situations.
It is possible to ascertain how − even in the most sophisticated scores for the
silents − repetitions and formulas remain relevant features: This is the
stronger link between the written practice of silent film music and
improvisation, a link that involves the musical material at a deeper level.
This independence of silent film music from the on-screen action due to live
performance and improvisational features was embedded in the discourse of
other early silent film music theorists. Many of the discussions reaffirmed
that, at the core, music for silent films is not something that plays with the
film, but, in a sense, next to the film. However, together with this largely
diffused opinion, an acknowledgement that silent film music could play a
more active part in the cinematographic communication started to spread.
The Italian critic and theorist Sebastiano Arturo Luciani wrote in 1919:
The music that was played during silent film screenings
was of two kinds: it included suites of marches and
dances, or pieces with a dramatic lyricism, usually taken
from famous operas. The first kind had a rhythmic
function, the second one was expressive. The two kinds
were alternatively used, whether the action had a
dynamic or pathetic character, that is to say whether the
music should unify a group of scenes or should express a
particular feeling in a certain scene. (Luciani 1980
[1919], 356; my translation)
Kieler Beiträge zur Filmmusikforschung, 9, 2013 // 53
Luciani recognized at least two active relations between music and moving
images. However, he added:
All the compositions especially written for cinema
screening still have not had any success and their role
was not different from the one of the music improvised
by the anonymous pianist we usually listen to: to fill the
void that the silent vision would produce. (Luciani 1980
[1919], 357)
This idea is reflected in Luciani’s definition of the so-called ›rhythmic‹
function: He does not say that the music follows the visual rhythms of the
film, but rather that it unifies a group of scenes. He was probably imagining
audiovisual situations in which a series of similarly paced actions share
common musical ideas conveying a similar pace, without precise
connections with the details of what was happening on screen. However,
even if Luciani considered this feature as a hindrance to cinema, it is evident
that he implicitly admits how music could enhance the perception of a film,
at least by underlining a relevant visual characteristic (the rhythm) and
creating a stronger sense of unification for a section of the spectacle. The
result is already an active audiovisual relationship, one in which music is
not just an adjunct to the image, but a contributive element.
Other voices in the early debates about film music agreed with Luciani in
testifying how even the most generic musical choices could benefit silent
film presentations, often by aiding the unity of the visual discourse. That
happened also in respect to the other audiovisual function identified by
Luciani, the ›expressive one‹. The way in which the music conveyed
Kieler Beiträge zur Filmmusikforschung, 9, 2013 // 54
consistent moods and emotions through different situations on screen could
also serve as a means of expressive unification. The 1920 manual Musical
Accompaniment of Moving Pictures by Edith Lang and George West
observed how the nature of the spectacle (based on performance and with
improvised features) would not allow constant close correspondence
between image and sound. The music was required to be ›generic‹, and to
use this necessity to become a unifier of different visual elements. Even if
the authors recommended a »close and minute following of every phase of
the photo play« (Lang/West 1920, 5), they had to admit that »[m]usic, it
may as well be stated, cannot always shift as quickly as will the facial play
of the actor in some scene or other. It will then behoove the player to give
the keynote of the situation with illustrative strains« (Lang/West 1920, 5).
This quotation introduced a key word in silent film music practice:
illustration. As it was used in one of the most authoritative historical texts
about silent film music theory and practice, the 1927 Allgemeines
Handbuch der Filmmusik by Hans Erdmann, Giuseppe Becce and Ludwig
Brav, it perfectly identified the two conflicting sides of this art. To illustrate
a film with music meant a dependence of the music on the meanings of the
images. However, it also meant that, within this dependence, the music was
asked just to draw an overall sketch of the rhythmic or expressive content of
the images. This is demonstrated by the structure of the Allgemeines
Handbuch der Filmmusik itself, which subdivides its repertoire into
categories marked by generic labels alluding to moods or situations, which
could occur in any film. This choice guaranteed a maximum of versatility to
the Handbuch. However, it also allowed musicians to stick with generic
image-sound relationships, with no urge to introduce more detailed and
specific interactions. It might be possible to describe the musical illustration
in this way: If the precision of some audiovisual relationships in sound
Kieler Beiträge zur Filmmusikforschung, 9, 2013 // 55
cinema can be compared to the accurateness of a still life or a portrait, the
overall audiovisual bond in silent cinema has frequently the expressive
freedom of a drawing done by memory. This is a key distinction between the
aesthetic of silent film music and the one of music for sound films; however,
this is also the aspect of silent film music that was most misunderstood. As
Ennio Simeon remarked, even Erdmann, Becce and Brav were induced to
overlook the creative potential of the audiovisual setting of silent film
because of a conception of the musical illustration that was too rigid. They
believed that because of the unavailability of a mechanical and fixed musicimage synchronization, even written scores could just hope for a role as
»author’s illustration« (Becce/Brav/Erdmann 1927, I, 6). Because of that,
Simeon commented:
The limit of Erdmann and Becce’s setting is […] never
going beyond the basic assumption which says that silent
film music has always and only to illustrate: the
prohibition, even at a theoretical level, of an active and
interactive role of the sound element is the proof of a
belief in a subordination of music to film and of a
confinement of it in a role of subsidiary art, a belief
which would not fail to retain an influence for decades
and that it is still not completely gone (Simeon, 1987, 76;
my translation).
Instead, the ›freedom‹ of the musical element in silent cinema does not
imply its complete passivity. The fact that the music can relate to the film by
freely spreading over the images without too strict obligations to the
Kieler Beiträge zur Filmmusikforschung, 9, 2013 // 56
rhythms and the moods is an element of peculiar richness. French critic
Arthur Hoérée wrote in 1934 a few meaningful considerations about this
topic and imagined an example:
Music is based on continuity, on the developments of
themes, following rules of its own. The film, on the
contrary, has to continuously break this continuity […]. It
works by contrast, while the music works by extension.
[…] Let’s imagine an episode where on one side there is
an aviator caught in a storm and on another side, as to
underline the contrast, there is his family happily
preparing for his return. If the director opposes eight
times these elements, as to stress the pathetic content of
the situation, the music should alternatively declaim The
King of the Elves by Schubert and murmur the Pastoral
Symphony by Beethoven… It is evident how it is
necessary to seek a mixed solution […]. It should be
sufficient to conceive a piece which is generally agitated
[…] in order to comment upon the struggle between the
aviator and the raging elements. The violins, however,
could sing a happy theme combined with the rest of the
polyphony. Such a music could accompany the two series
of sights […] and combine them in a sort of synthesis.
The simultaneity of the sound is here more correct than
the film, as the two groups of images show simultaneous
situations. So the music fixes the conventional character
of the film, it completes it or, better, it explains it
(Hoérée, 1934, 46-47; my translation).
Kieler Beiträge zur Filmmusikforschung, 9, 2013 // 57
By following these suggestions, it is possible to argue that when the lack of
tight bonds between music and the moving image was used with creative
purpose, it paved the way for effective audiovisual settings that sound film
could not exactly replicate because of its different audiovisual nature (e. g.
because of the creation of an expectation in the audience for diegetic or
nondiegetic sounds, using Claudia Gorbman’s terminology [Gorbman 1987,
22-26]: a kind of expectation which was not part of the experience of the
silent film moviegoer).
2. The ›bridge‹ function
It seems suitable to refer to the first audiovisual relationship discussed here
as ›bridge function‹. It is directly related to the capability of silent film
music to freely expand over different sequences. At the origin of the
›bridge‹ function there surely are musical routines related to improvisation,
and to improvisation done with a lack of attention or accuracy in particular.
Before becoming a conscious audiovisual strategy, the ›bridge‹ function
could often have been the result of the work of a lazy or tired pianist, who
kept repeating a certain musical mood or episode without concern for the
images on the screen.
In fact, evidence of the presence of a ›bridge‹ function in cue sheets,
compilations or full scores is actually the repetition. The prevalent
dramaturgical model appears as follows: a certain visual or narrative
element, with some form of relevance, triggers a musical episode which is
pertinent to the content of the scene. For instance, if there is a sad scene, a
sad music starts. At the beginning, the audiovisual setting is quite
Kieler Beiträge zur Filmmusikforschung, 9, 2013 // 58
conventional. It could be identified with one of the two basic functions that
Sergio Miceli identifies in film music as a whole, that is to say:
accompaniment and commentary (Miceli 2009, 632-635), which more or
less coincide with the two options suggested by Luciani, the »rhythmic
function« and the »expressive function«. As the film continues, however, the
on-screen action distances itself from the musical element. The music keeps
repeating the same idea (a theme, or also a little piece with a simple
structure), with no or little developments or variations. The monotony of the
sound retains thus a memory of the visual element from which the
repetitions started, until a new relevant feature of the film requires a change
in the music.
Many pertinent examples of that could be quoted from Hugo Riesenfield’s
illustration for CARMEN (USA 1915, Cecil B. DeMille), reconstructed in
1991 by Gillian B. Anderson and based on music by Georges Bizet
(Anderson 2005).6 One instance is the sequence of the fight between José
and lieutenant Zuniga, a sequence that culminates in the killing of the latter.
When José and Carmen enter a tavern, a place where they are going to meet
the smugglers and Zuniga, the music introduces the Danse bohémienne from
the Suite La jolie fille de Perth. The apparent function is the one of a
comment (or an expressive function): the graceful mood of the melody is
connected with the romantic undertones of the entrance of José and Carmen
together. These undertones will be the cause of Zuniga’s scorn and of the
6
The musical material of this illustration mainly comes from the two symphonic
suites from the opera by Georges Bizet. However, there are also some arias
(L’amour est enfant de Bohème, La fleur que vous m’avez donnée ), which Anderson
included because some documents testify the presence of singers during the first
screenings of the film. However, there are also quotations from other works by
Bizet, like the two Suites from L’Arlésienne and the Suite from the opera La jolie
fille de Perth.
Kieler Beiträge zur Filmmusikforschung, 9, 2013 // 59
following fight: notwithstanding this, the violence and the pace of the
events, let alone the sudden mutation of the mood, do not compel the music
to change. Instead, the score asks for an almost obsessive repetition of a
long episode from the Danse, featuring a progressive enrichment of the
orchestration and a modulation from B minor to F sharp major, two
elements whose potential connection with the scene (they could vaguely
hint at the increase of tension during the fight) is negated by the repetition
that restores the initial aspect of the music. When the action leaves the
tavern, the music comes to an end: the apparition of a shot of Carmen
approaching a river is then commented by the famous Habanera from the
original opera. So, the whole sequence appears ›bridged‹ together by a
music that constantly reminds the audience of the reason for the fight, more
than commenting upon the fight itself.
It is interesting to note how in CARMEN some of the elements that most
frequently bring a ›bridge‹ function to an end are intertitles with a
descriptive purpose, signaling a shift from one location to another. In the
principal document that Anderson used for her reconstruction − the piano
score from Riesenfeld’s illustration preserved by the Library of Congress −
the music is often accompanied by cues that explicitly ask for the repetition
of a certain episode until the apparition of a certain intertitle, using the
formula: »Play until Title« (Anderson 2005, 30).
Not every use of the ›bridge‹ function is an effective one. It is in fact easy
for this function to lose its meaning and return to its origin of passive and
inaccurate repetition. The reason of that lies in the fact that the ›bridge‹
function can be recognized only if framed between two other audiovisual
functions. It is possible to argue that a ›bridge‹ function can only be defined
by the presence and the relationship with other audiovisual settings. If the
Kieler Beiträge zur Filmmusikforschung, 9, 2013 // 60
beginning and the ending of the ›bridge‹ do not convey a sufficient
audiovisual meaning, the purpose of the ›bridge‹ itself becomes
inconsistent. This is what differentiates the ›bridge‹ function of silent
cinema from the ›continuity‹ function described by Gorbman and others in
their discussion of sound films (Gorbman 1987, 25-26). The continuity
function has always an explainable narrative purpose and it can use the
diegetic/nondiegetic dialectic. It can ›bridge‹ together »two spatially
discontinuous shots«, or it can serve as a »depth cue« by letting the same
diegetic music play louder or softer according to the different supposed
distance between the spectator and the sound source. But those are just two
basic examples from a wide set of occurrences, where the continuity
function is always »a nonrepresentational provider of relations, among all
levels of the narration«. (Gorbman 1987, 26). Instead, the silent film
›bridge‹ function is not related to the narrative. It can sometimes have an
impact on the narration, thus creating a continuity function, but this is a side
effect. At its core, the ›bridge‹ function is a connecting tissue between two
other audiovisual functions, which occur at a relevant temporal distance
from each other. So, it is a way to fill the gap between two meaningful
associations between image and music, disregarding the actual narrative
content of the scenes framed by those functions. More than a function, this
is a configuration of functions. It is necessary to keep in mind this when
analyzing silent film music, in order to avoid confusion between this weak
and relative, yet meaningful configuration, and mere cases of unconsidered
and accidental repetitions. In sound cinema, the bridge function could
theoretically be used, but it actually lost its purpose, as it became possible to
stop the music and continue the audiovisual interactions with other sound
configuration, or to create a shift between the diegetic and nondiegetic
levels of sound.
Kieler Beiträge zur Filmmusikforschung, 9, 2013 // 61
3. The ›interdiegetic‹ function
A second instance of an audiovisual strategy that is specific to silent film
can be identified by considering precisely the relationship that exists
between the music and the narrative world of the film: the diegesis (Genette
1976 [1972], 75; Genette 1987 [1983], 12). In order to develop this topic, it
is first of all necessary to recall the system of »levels« of film music, as
explained by Sergio Miceli.
Miceli identified three levels of interaction between music and film
narrative. The first one is the internal level: the music is »produced in the
narrative context of the scene/sequence« (Miceli 2009, 643). It corresponds
to an audiovisual setting that is called »diegetic music« by other authors, for
example Claudia Gorbman (1987, 22-26) or David Bordwell and Kristin
Thompson (2003 [1979]). Basically, it is possible to say that every
intervention of the music that can be perceived by both the audience and the
characters on-screen belongs to the internal level.
The external level proposes a setting where the music is audible by the
audience, but not by the characters (Miceli 2009, 649). Miceli argued that
there could be two occurrences of this level: a non-critical one and a critical
one. The non-critical external level is identified when »a musical event […]
just confirms and reinforces the expressive content of the episode« (Miceli,
2009 651). The critical one, instead, »comments upon the episode of the
film by using contrasting ideas and by generating a semantic short circuit
which denies the expectations of the spectator and asks for an active role,
that is to say an interpretative role« (Miceli 2009, 652). Even if it is possible
to conceive a critical external level in silent film music, it seems that it did
not belong to the historical practice. The illustrative basis of silent film
Kieler Beiträge zur Filmmusikforschung, 9, 2013 // 62
music usually asked for parallelism and mutual reinforcement of music and
image. Counter-examples were rather individual incidents than expressive
ideas: as Stephen Bottomore (1995, 120) recalled, »sometimes [the pianists]
misjudged the mood badly, and one correspondent complained in 1912: ›I
find nothing more irritating than to have to listen to variations on Ginger,
you’re barmy when the operator is showing a serious dramatic film‹«.
Miceli also talks about a mediate level. This includes all the instances where
the music is accessible within the diegesis, but is not equally audible by all
the characters. It includes music perceived through memory of a previous
musical event on the internal level, that is, so to say, shared with the
audience (Miceli 2009, 654-657). Miceli uses this classification to speak
about film music in general; however, it seems difficult to completely adapt
it to silent film music. In fact, one of these levels cannot be found in silent
cinema: the internal level.
The central point of this problem lies within what Zofia Lissa already
observed in 1965: »In silent film, the music […] was an instrument
connected with the image itself in a purely external way« (99-100; my
translation). In sound cinema, the definitive and unchangeable recording of
an acoustic event in synchrony with a visual event generates the illusion of
the internal level. The sound ›belongs‹ to the images because of a threefold
bind: qualitative, temporal and mechanical. In fact, the link between the two
elements does not depend only on the quality and the intensity of the sound,
which is consistent with the expectations of the audience in relation to a
certain visual feature, or by the simultaneous occurrence. The two elements
are also materially related, because of their respective positions on the
soundtrack and on the visual track of the film. This threefold bind makes the
audiovisual relationship an obligatory one. In silent film, instead, the
Kieler Beiträge zur Filmmusikforschung, 9, 2013 // 63
mechanical association is not a rule: usually, the film contains only a visual
track (the sequence of frames). The qualitative association is moreover used
with great freedom: a visual event could be associated with sounds with an
expressive meaning, which can be achieved even if those sound effects or
musical elements are not perfect simulations of sounds heard in everyday
life. Finally, the synchrony could be used also in silent cinema, but it is not
compulsory. For example, the image of a firing cannon could be
accompanied by absolute silence: If this choice is adequately prepared and it
is pertinent with the sense of that particular sequence, the audience would
accept it as an expressive feature and not as if something is ›missing‹. This
could of course also happen in sound cinema; however, in sound cinema the
silence would be much more difficult to accept. This is because of the
implicit agreement between the director and the audience. In sound cinema,
the audience expects to be able to hear the sounds that come from the
narrative world; the inaccessibility is an exception. The central reason
behind this is the fact that in sound cinema, the most important acoustic
element is the human voice. As Michel Chion said, the cinema after 1927
became vococentric: the voice of the actor, which comes from the world of
the film, must be constantly perceivable (Chion 1982, 15). In silent cinema,
the exact opposite is true: the audience does not expect to be able to hear the
sounds coming from within the film. While in sound cinema the triple bind
(qualitative, temporal and mechanical) is usually an obligatory one, in silent
cinema it is a choice.
Kieler Beiträge zur Filmmusikforschung, 9, 2013 // 64
On top of that, in silent cinema there is no fixed hierarchy of importance
between acoustic events. The human voice, when used, is just a sound
among the others. It could become extremely relevant in practices of silent
film accompaniment based on lecturers and readers (Altman 2004, 55-72) or
on more refined performances of actors, as in the case of the Japanese
benshi. It was not, however, the dominant sound by rule.
It is possible to comment upon this by considering a short sequence from the
film THE CAT
AND THE
CANARY (USA 1927, Paul Leni) and from its
musical score written by British composer Neil Brand in 2004. The use of
examples from contemporary authors is justified here by the purpose of the
present article, which does not intend to develop a discourse about the
change in the reception of audiovisual strategies from historical audiences to
the present day's public. The intent here is instead to point out how both
historical and contemporary musical illustrations of silents take equally
advantage of the absence of diegetic sound to propose audiovisual settings
based upon an apparent diegetic ambiguity.
In the film by Leni, while the lawyer and the housemaid Mammy Pleasant
are arguing inside a house, the detail of an unknown hand knocking at the
door is superimposed to the frame. The two characters react to this, so it is
implied that they heard the knocking. The audience, on the other hand, is not
supposed to hear anything. A causal relationship between the events derives
from the visual superimposition alone. Neil Brand, however, decided to join
the image of the hand knocking with the rhythmic sound of a kettle drum.
That was an expressive choice, evidently aimed to reinforce the already
evident meaning of the images. Even without the synchrony with the kettle
drum, the sequence would have been perfectly understood by the spectator.
In sound cinema, instead, the silence or the absence of synchrony would
Kieler Beiträge zur Filmmusikforschung, 9, 2013 // 65
have added a sense of strangeness and unreality to the sequence. On the
contrary, the presence of this synchrony in a silent film, precisely because it
was not strictly necessary, becomes a way to emphasize the expressive
meaning of the image: in this case, the image of the hand becomes
particularly ominous and unsettling.
Even with such a short example, it is possible to understand how silent
cinema learnt to take advantage of the absence of a diegetic level of the
sound. One of the most interesting results of this premise is a function that
could be called ›interdiegetic‹, or ›interlevel‹ function. It is a function that
uses the freedom of association between music and image in silent cinema
to create situations in which the music is simultaneously extraneous and in
relation to the diegesis, in a way different from sound cinema because of the
different basis of its communicative agreement with the audience. The term
›metadiegetic‹ could be used as well; however, it has already come into use
in film music analysis through Gorbman’s work on narrative film music,
with the sense of a sound event expressing subjective perceptions (Gorbman
1987, 26). In the present article, the meaning of ›metadiegetic‹ would
instead be closer to its original sense in Gérard Genette’s theory of
narrativity, where it identifies a situation in which a character narrates a
story within the main story, creating a frame where different levels of
diegesis are convoluted (Genette 1976 [1972], 276). To avoid confusion, the
term ›interdiegetic‹ will be used.
Kieler Beiträge zur Filmmusikforschung, 9, 2013 // 66
The basic setting of the interdiegetic function is the following: the music is
pertinent to more than one of the three levels identified by Miceli at the
same time. Each one of these levels is equally related with the music: there
is no predominant level. Also, within each level it could relate
simultaneously to multiple visual features, without creating distraction or
awkwardness in the spectator.
As an example, it could be useful to quote a fragment from Detlev Glanert’s
reconstruction of Giuseppe Becce’s illustration for DER LETZTE MANN
(Germany 1924, Friedrich Wilhelm Murnau), which was published on DVD
by Transit Film in 2003. After having lost his job as chief doorman, Emil
Jannings returns to his house, drunk. While his neighbor, the woman who
will spread the news of the old man’s misfortune, enters his apartment,
Jannings takes his whistle, one of the symbols of his former role, and blows
it. This event is synchronized with a high-pitched trill of a solo piccolo flute.
This trill is simultaneously part of two levels: it hints to the internal level (it
imitates the sound of the whistle), but it is also part of a musical discourse
of external level: a discourse that precedes and follows the trill, and which is
logically and aesthetically connected with it. The music, before the trill, is
reaffirming the fifth grade of the tonality of C major and stops on a chord
(G-B-D), which has a D as its highest note. The trill is, coherently, based on
a D. After the trill, the music makes a brief cadenza and regularly moves to
C major. But this sequence is also noteworthy because of the ambivalence of
the music within the apparent internal level: thanks to the editing, which
alternates a shot of Jannings with the image of the neighbor laughing, the
trill feels like a simultaneous allusion to the whistle and to the shrill laughter
of the woman. Cause and effect of two actions collapse into a single
acoustic event and join in the grotesque, the one who is laughable and the
one who laughs, believing to be superior.
Kieler Beiträge zur Filmmusikforschung, 9, 2013 // 67
The sequence continues with the sound of the piccolo, imitating the whistle
and the laughter, mixed with the orchestral accompaniment. Soon after that,
Murnau introduces a new event with an important musical potential. Two
drunk men stop under the window of Jannings’ apartment, and one of them
starts playing a trumpet. The man is interrupted twice by his companion. In
these instances, Glanert’s reconstruction uses a close synchronization with
the pace of the action: we can hear the sound of a solo trumpet just when we
see the character playing it. Also, the way the sound is intonated and the
way in which it stops are quite consistent with the situation that is being
enacted. This really seems a simulation of an internal level; however, the
illusion is negated by the fact that we are not able to hear the voices of the
two men, which visibly start to speak as soon as the trumpet player stops.
The partial allusion to the internal level rendered by the music, in the silent
context, stands out as an expressive choice; in a sound film, instead, it
would have been the absence of the voices which would have unsettled the
spectator the most. Moreover, in case of a real silent film screening, the
audience would be constantly conscious of the presence of a real trumpet
player in the cinema hall.
When finally the man with the trumpet manages to start playing a full piece,
the sound of an orchestra appears as a discreet accompaniment to the main
melody. The behavior of Jannings at the window clearly communicates that
he is actually listening to the sound of the trumpet: however, the musical
dramaturgy also implies that he could not be listening to the same sound the
audience is hearing, because in the world of the film there could not be an
orchestra at that moment and in that place. In a sound film, this would have
been a solution hinting maybe to a mediate level (Jannings hears a trumpet
melody which the audience hears as well, but his imagination »completes«
it with the sound of an orchestra). In a silent film, it is not possible to clearly
Kieler Beiträge zur Filmmusikforschung, 9, 2013 // 68
classify the level of this musical episode, because we cannot know if the
melody Jannings is hearing is exactly the same melody which is being
played in the world of the film. So, it could certainly be a music staying in
the mediate level, but it could also be a music from the external level, at the
same time.
Another instance can be found in a musical illustration composed in 1992
by Richard McLaughlin for Jean Grémillon’s GARDIENS
DE PHARE
(France
1929).7 The film deals with a lighthouse guardian and his son, who have to
live isolated on an island for a long time. However, the son has been
infected with rabies by a dog bite just before their isolation begins. He
slowly slips into a feverish state which leads him to bursts of violence and
hallucinations.
A sequence of the film shows the old guardian remembering a country
festival. The audience can see a scene where young men and women dance
while some musicians play violins and drums. McLaughlin coherently
decided to use the instruments suggested by the images to illustrate the
situation. However, it is not possible to say that this music is part of the
internal level. Even if the timbres of the instruments are correct, in fact,
there is no synchrony between the music and the image: the blows from the
drums do not follow the movements on screen, as well as it is clear how the
violins in the film are not playing what the audience can hear. So, the music
is separated from the diegesis, but it is related with it at the same time.
However, the music also belongs to the mediate level. In fact, it appears
inside a memory of the old guardian, and it appropriately continues when
7
This illustration was presented at the International Silent Film Festival Le Giornate
del Cinema Muto on October 18, 1992. It was recorded on a videotape which is
property of the Cineteca del Friuli in Gemona, Italy.
Kieler Beiträge zur Filmmusikforschung, 9, 2013 // 69
the film reintroduces the character in his present situation, at the lighthouse.
So, the music can pass simultaneously through the various levels,
›embracing‹ them from a privileged position. It becomes in a sense, an
ubiquitous entity.
This ubiquity of the music is different from the kind that can be found in
sound cinema, because of the already cited ›agreement‹ between the director
and the audience: the inner acoustic world of the film must always be
accessible. An example of an attempt of ›ubiquity‹ in sound cinema
mentioned by Miceli (2009, 662-663) involves a sequence from Milos
Forman’s AMADEUS (1984): as Mozart dictates his Requiem to Antonio
Salieri, we hear the materialization of his creative thoughts in the form of a
full orchestral rendering. This appears as a mediate level, which, however,
seems to suddenly shift to an external level as the director cuts to the coach
who is bringing back Mozart’s wife to his dying husband. But actually, the
music is always positioned in a mediate level by the constant presence of an
element from the internal level. The feeble singing of the ill Mozart, which
the orchestra transforms into the finished Requiem, clarifies the role of the
music proposed by soundtrack, making it a projection of the composer’s
mind. In a silent film, there would not have been such a reference. The
music would always have belonged to more than one level simultaneously:
in this case possibly the mediate and the external one.
4. Conclusion
The ›bridge‹ and the ›interdiegetic‹ functions have been described as to
suggest an integration to the usual schemes of analysis of film music, in
Kieler Beiträge zur Filmmusikforschung, 9, 2013 // 70
order to provide theoretical devices which would be more fitting to the
necessities of the audiovisual language of silent cinema. This kind of
suggestion aims to integrate the already existing theoretical schemes, and
not to substitute them or to deny their effectiveness. In fact, one of the most
urgent problems film music theory has to face is surely the absence of a
dialogue between the many languages of film music analysis, because of a
proliferation of classifications of functions which often overlap. This is a
problem that was evidenced as early as 1981 by Hansjörg Pauli, who
criticized Lissa’s description of eleven audiovisual functions calling it the
consequence of a »Systematisierungswut« (Pauli 1981, 187), that is to say a
»fury in systematizing«. The ›bridge‹ and ›interdiegetic‹ functions, however,
should not be regarded just as two more theoretical categories. Instead, they
are material indicators of the nature of the audiovisual setting of silent
cinema. In this regard, and as a final consideration, it could be said that a
description of the audiovisual functions of silent cinema might lead towards
a definition of silent cinema as a whole. It seems in fact difficult to find
relevant differences between silent and sound cinema, if the comparison is
made just on the basis of the visual language. Early cinema certainly
displayed some peculiar visual traits (like the fixed camera and the theatrical
setting) that make it immediately recognizable. The presence of the
intertitles is as well often considered a typical feature of silent cinema.
However, these elements are not indispensable. In fact, it is possible to
conceive a sound film with theatrical setting and fixed camera; also, there
have been silent films, which used only a few or no intertitles (DER LETZTE
MANN was among them). On the other hand, it is as well possible to
imagine a silent film created with the contemporary cinema technologies or
aesthetics: it might be sufficient to cite Aki Kaurismäki’s JUHA (Finland
1999), which had a recorded soundtrack, but was nonetheless silent mostly
Kieler Beiträge zur Filmmusikforschung, 9, 2013 // 71
because of his use of the human voice as a sound among sounds, with no
central role, and because of the presence of a score by Anssi Tikanmäki
which was prepared to be performed live during the film screening.
Therefore, it could be suggested here that, more than the images, it was the
presentation of sound and music, which identified the true nature of silent
cinema as a communicative device. The performance and improvisation
(which are the causes behind the ›bridge‹ function) gave a distinct
uniqueness to each screening. Silent films imposed conditions to the cinema
hall that were similar to those of a concert hall which, as Tomlinson Holman
argued, »is a space for production […] [while] a movie theater is a space for
reproduction« (Lo Brutto 1994, 204). In silent cinema, every new screening
could be legitimately different from the precedent and the differences could
also be dramatic, especially in the case of improvisations. Along with that,
the ›interdiegetic‹ function reveals how the music and the sound did not
have to comply any obligation towards a fixed acoustic hierarchy with the
human voice at the top, and also how the sound could fulfill the
expectations of the audience without adhering to a strict division between
the diegetic levels or to a synchrony mimicking the perception of sound
experiences in everyday life. Even without trying to propose a final
definition of silent cinema, which would go beyond the scope of this work,
the study of the ›bridge‹ and ›interdiegetic‹ function could at least invite to
confirm, once more, that silent cinema really was not a precursor or a less
developed ancestor to sound cinema. It was, and still is, a different instance
of audiovisual entertainment: another cinema.
Kieler Beiträge zur Filmmusikforschung, 9, 2013 // 72
Bibliography
Alperson, Philip (1984) On Musical Improvisation. In: The Journal of Aesthetics
and Art Criticism 43:1, pp. 17-29.
Altman, Rick (2004) Silent Film Sound, New York & Chichester: Columbia
University Press.
Anderson, Gillian (1988), Music for Silent Films 1894-1929: a Guide, Washington:
Library of Congress.
______________ (2004) Musical Missionaries: ›Suitable‹ Music in the Cinema
1913-1915. In: Miceli, Sergio (ed.) (2004), La musica nel cinema. Tematiche e
metodi di ricerca (Civiltà musicale no. 51/52).
______________ (2005) Geraldine Farrar and Cecil B. DeMille: The Effect of
Opera on Film and Film on Opera. In: Perriam, Christopher; Davis, Ann (eds.)
(2005), Carmen: From Silent Film to MTV, Amsterdam & New York: Rodopi,
pp. 23-35.
Balázs, Béla (1948) Der Film. Werden und Wesen einer neuen Kunst, Wien: Glibus
Verlag, Italian translation by Grazia and Fernaldo Di Giammatteo, Grazia; Di
Giammatteo, Fernaldo (1975) as Il film. Evoluzione ed essenza di un’arte
nuova, Torino: Einaudi.
Becce, Giuseppe; Brav, Ludwig; Erdmann, Hans (1927) Allgemeines Handbuch
der Filmmusik, 2 vols., Berlin: Schlesinger´sche Buchhandlung.
Beynon, George (1921) Musical Presentation of Motion Pictures, New York:
Schirmer.
Bordwell, David; Thompson, Kristin (2003 [1979]) Film Art: An Introduction,
New York, McGraw-Hill.
Bottomore, Stephen (1995) I Want to See this Annie Mattygraph/Voglio vedere
quest’Annie Matografo, Pordenone: Le Giornate del Cinema Muto.
Cavalcanti, Alberto (1939) Sound in Films. In: Weis, Elisabeth; Welton, John (eds.)
(1985) Film Sound. Theory and Practice , New York: Columbia University
Press, pp. 98-111.
Chion, Michel (1982) La voix au cinema, Paris: Éditions de l'Étoile.
Hampton, Benjamin (1970 [1931]) History of the American Film Industry, New
York: Dover.
Herzog, Charlotte (1981) The Movie Palace and the Theatrical Sources of Its
Architectural Style. In: Cinema Journal 20: 2, pp. 15-37.
Kieler Beiträge zur Filmmusikforschung, 9, 2013 // 73
Brand, Neil (1998) Dramatic Notes. Foregrounding Music in the Dramatic
Experience, Luton: University of Luton Press.
Brownlow, Kevin (1981) Abel Gance’s Napoléon Returns from Exile. In:
American Film 6, pp. 28-32, 68, 70-73.
Genette, Gérard (1976 [1972]) Figure III. Il discorso del racconto , Italian trans. by
Lina Zecchi, Torino: Einaudi.
_____________ (1987 [1983]), Nuovo discorso del racconto, Italian trans. by Lina
Zecchi, Torino: Einaudi.
Gorbman, Claudia (1987) Unheard Melodies. Narrative Film Music , Bloomington
& Indianapolis: Indiana University Press.
Gunning, Tom (1996) »Now You See It, Now You Don’t«: The Temporality of the
Cinema of Attractions. In: Abel, Richard (ed.) (1996), Silent Film, London:
Athlone Press, pp. 71-84.
Hoérée, Arthur (1934) Essai d’esthétique du sonore. In: La revue musicale 151, pp.
45-62.
Holman, Roger; Gaudreault, André (eds.) (1982) Cinema 1900-1906: An
Analytical Study, 2 vols., Brussels: FIAF.
Lang, Edith; West, George (1920) Musical Accompaniment of Moving Pictures ,
New York: Schirmer.
Lissa, Zofia (1965) Ästhetik der Filmmusik, Berlin: Henschelverlag.
Lo Brutto, Vincent (1994) Sound-on-Film: Interviews with Creators of Film Sound ,
Westport: Praeger Publishers.
London, Kurt (1970 [1936]) Film Music. A Summary of the Characteristic features
of its History, Aesthetics, Technique; and possible Developments, New York:
Arno Press.
Luciani, Sebastiano Arturo (1919) La musica al cinematografo. In: Il Primato
Artistico Italiano, vol. I, no. 2, reprinted in (1980) Tra una film e l’altra.
Materiali sul cinema italiano 1907-1920, Venezia: Marsilio, 1980.
Miceli, Sergio (2000) Musica e cinema nella cultura del Novecento , Milano:
Sansoni.
___________ (2009) Musica per film. Storia, Estetica, Analisi , Tipologie, Milano:
Ricordi.
Pauli, Hansjörg (1981), Filmmusik: Stummfilm, Stuttgart: Klett Cotta.
Redi, Riccardo (1999) Cinema muto italiano (1896-1930), Roma: Biblioteca di
Bianco e Nero.
Kieler Beiträge zur Filmmusikforschung, 9, 2013 // 74
Simeon, Ennio (1987) Appunti sulla teoria e prassi musicale nel cinema muto
Tedesco. In: Trento Cinema. Incontri Internazionali con il film d’autore , Trento:
Servizio Attività Culturali della Provincia Autonoma di Trento, pp. 74-80.
____________(1995) Per un pugno di note. Storia, teoria, estetica della musica per
il cinema, la televisione e il video, Milano: Rugginenti.
Kieler Beiträge zur Filmmusikforschung, 9, 2013 // 75
Empfohlene Zitierweise
Bellano, Marco: Silent Strategies: Audiovisual Functions of the Music for Silent Cinema.
In: Kieler Beiträge zur Filmmusikforschung 9, 2013, S. 46-76.
URL: http://www.filmmusik.uni-kiel.de/KB9/KB9-Bellano.pdf
Datum des Zugriffs: 28.2.2013.
Kieler Beiträge zur Filmmusikforschung (ISSN 1866-4768)
Copyright © für diesen Artikel by Marco Bellano.
All rights reserved.
Copyright © für diese Ausgabe by Kieler Gesellschaft für Filmmusikforschung.
All rights reserved.
This work may be copied for non-profit educational use if proper credit is given to the
author and Kieler Beiträge zur Filmmusikforschung.
Kieler Beiträge zur Filmmusikforschung, 9, 2013 // 76