4 - Stage and Screen

Cambridge Books Online
http://ebooks.cambridge.org/
A History of Film Music
Mervyn Cooke
Book DOI: http://dx.doi.org/10.1017/CBO9780511814341
Online ISBN: 9780511814341
Hardback ISBN: 9780521811736
Paperback ISBN: 9780521010481
Chapter
4 - Stage and screen pp. 131-182
Chapter DOI: http://dx.doi.org/10.1017/CBO9780511814341.005
Cambridge University Press

4 Stage and screen
In his short but provocative discussion of the pros and cons of filming
opera, Béla Balázs drew a clear distinction between the straightforward filmic
preservation of theatrical opera productions (which he viewed as ‘very useful
in improving the musical taste of the public’) and the exciting possibility
of a film opera ‘intended and directed and composed for the film from the
start, [being] a new musical form of art with new problems and new tasks’
(Balázs 1953, 275). Echoing the opinions of commentators in the 1930s,
he dismissed the notion of filming pre-existing stage works in a realistic
cinematic style as completely incompatible with the highly stylized idiom
of operatic acting and singing. Balázs nevertheless advocated a degree of
flexibility of direction and, above all, mobile camerawork in order ‘to loosen
up the old-fashioned rigidity which is scarcely tolerable even on the stage
to-day’, and praised René Clair for his ability to parody in effective cinematic
terms the ‘grotesquely unnatural character of stage style’ by such fluidity of
directorial technique (Balázs 1953, 276). Balázs’s mention in this context of
Clair, whose bold and imaginative ideal of the ‘musical film’ was examined
in Chapter 2, suggests that certain fundamental aesthetic considerations
inform an understanding of the pleasures and pitfalls of reworking both
highbrow and middlebrow music-oriented dramatic forms on celluloid.
Although Clair’s ‘musical film’ quickly gave way to the more commercially
minded ‘film musical’, these parallels persisted, especially when film musicals
were based directly on stage works.
This chapter will briefly summarize trends in opera films from the silent
era to modern times, and the role of operatic music and settings in other
types of film, before tracing the development of the far more populist,
prodigious and lucrative film musical in Hollywood and elsewhere. (For
discussion of the musical in Indian cinema, see Chapter 9.) Issues of elitism
and prestige at the heart of filmed opera, very different from the sheer enter-
tainment and vernacular energy of the best film musicals, also feed into a
discussion of the role of film music in the significant body of sound films
based on Shakespeare’s plays, which deserve a focused examination not
only because they have frequently involved the skills of front-rank com-
posers specifically hired to lend the products a prestige similar to that of
opera, but also represent encouragingly varied and often genuinely creative
responses to similar problems of musico-dramatic style and structure – and
sometimes also spring from a desire to bring high culture to the masses.
[131]
Downloaded from Cambridge Books Online by IP 198.211.119.232 on Sat Apr 16 00:58:27 BST 2016.
http://dx.doi.org/10.1017/CBO9780511814341.005
Cambridge Books Online © Cambridge University Press, 2016
132 A history of film music
Opera on film
Cinematic interpretations of scenes from popular operas were widespread
in the era of the silent film. By as early as 1904, extracts from Wagner’s Parsi-
fal, Rossini’s Barber of Seville and Gounod’s Faust had been recorded on film
and projected with live music. The sumptuous Gounod film, Faust et Mar-
guerite, was directed by pioneering film-maker Georges Méliès, who him-
self appeared as Méphistophélès. In 1908, two developments contributed
towards a boom in such ventures: first, the founding of the influential
film d’art movement in France (see Chapter 1), which intensified inter-
est in cinema as an art form; second, the introduction of more stringent
copyright legislation which henceforth compelled film-makers to plunder
non-copyright classics of literature and the operatic stage in the interests of
economy. A further consideration in the issuing of operatic extracts on film
was the desire to promote gramophone recordings of the singers featured, a
commercial concern that sat uncomfortably alongside the growing feeling
that operatic source material could lend the medium of film a prestige that
had formerly eluded it.
‘Canned theatre’ (in the parlance of the day) thus came to include ‘canned
opera’, with both Pathé and Edison releasing new versions of Gounod’s Faust,
in 1909 and 1911 respectively; Pathé also made a film based on Verdi’s
Il trovatore in 1910. Other operas subjected to film treatment included
Thomas’s Mignon, Auber’s Fra Diavolo and Wagner’s Siegfried (all 1912).
Appropriate musical extracts from the operas were suggested by the distrib-
utors to guide projection venues in fitting live performances to the images,
and these performances generally lacked the vocal parts (Marks 1997, 72);
venues were at liberty to use any other music of their choice, and often
did. In the case of a film based on Mozart’s Le nozze di Figaro (1914), the
distributor declared in its trade advertisement that ‘Music adapted from
the famous Opera will be supplied gratis’ (Marks 1997, 193). Biopics based
on the lives of Wagner and Verdi appeared in Europe in 1913 and 1914
respectively, and operatic films continued to be produced in large numbers
in France, Germany and Italy. Bizet’s ever-popular Carmen, already given a
filmed treatment by Edison in 1910, formed the basis for a silent film star-
ring operatic soprano Geraldine Farrar in 1915 – and wickedly parodied
a year later by Chaplin in his Burlesque on Carmen. Produced by Cecil B.
DeMille for Paramount, the Farrar film was launched in the USA to the
accompaniment of arrangements from Bizet’s score prepared by Riesenfeld
and Rothapfel. In 1917, Puccini refused to allow his music to be used for
a film version of La Bohème, his gesture reinforcing the problem of tack-
ling works still in copyright; but production of non-copyright opera films
continued apace. By the 1920s silent films had grown longer in duration
133 Stage and screen
and relatively sophisticated in technique, and later operatic ventures were

accordingly more satisfying. Among them, several films based on La Bohème
finally appeared (in 1921, 1922 and 1926).
A silent film of Strauss’s Der Rosenkavalier was made in 1925, and this
is occasionally screened today – as it was at Aldeburgh in 2002 – with live
piano accompaniment. The film’s production credentials were impressive,
with direction by Robert Wiene (responsible for the legendary expressionist
film The Cabinet of Dr Caligari in 1919), design by Alfred Roller (designer of
Max Reinhardt’s original staging of the opera) and the direct participation
of both librettist and composer. Strauss originally intended to have little
to do with the project, which was scheduled to be screened at the Dresden
Opera House in January 1926; he agreed to compose a new march, but left
the task of adaptation to Otto Singer and Karl Alwin. In December 1925 the
librettist Hofmannsthal wrote to the composer to say: ‘I must admit that
your refusal to conduct the film in Dresden came quite unexpected and is a
grave blow to me . . . I cling to the hope that it may perhaps be tempered with
“blessed revocability”, but if it were irrevocable, I foresee for you (and conse-
quently also for me) . . . the loss of very considerable financial expectations’
(Hammelmann and Osers 1961, 411). Hofmannsthal was impecunious at
the time and clearly had a vested interest in generating revenue from the
project: his awareness of its commercial potential is telling. In a later letter
Hofmannsthal passionately argued that the film would provide ‘a positive
fillip and new impetus to the opera’s success in the theatre’ (quoted in Jef-
ferson 1985, 123). Strauss duly relented, and conducted the shambolic first
screening in which it rapidly became apparent that he possessed neither
the technical expertise nor sympathy with the medium necessary to ensure
accurate synchronization with the projected images: he suffered the humil-
iation of being forced to yield his baton to an experienced film conductor
(London 1936, 69). In April 1926 Strauss was in London to conduct the first
English screening, on which occasion he also recorded orchestral excerpts
with the Tivoli Orchestra. Released by His Master’s Voice, these recordings
were an early example of a film ‘tie-in’.
Reflecting on the tension between populist film music and modernist
music ‘driven into the esoteric’ because of its minority appeal, Eisler and
Adorno took a hearty swipe at both Der Rosenkavalier and its composer:
at the time when motion-picture music was in its rudimentary stage, the
breach between middle-class audiences and the really serious music which
expressed the situation of the middle classes had become unbridgeable. This
breach can be traced back as far as Tristan, a work that has probably never
been understood and liked as much as Aı̈da, Carmen, or even the
Meistersinger. The operatic theater became finally estranged from its
audience between 1900 and 1910, with the production of Salome and
Electra, the two advanced operas of Richard Strauss. The fact that after 1910,
with the Rosenkavalier – it is no accident that this opera has been made into
a moving picture – he turned to a retrospective stylized way of writing
reflects his awareness of that breach. Strauss was one of the first to attempt
to bridge the gap between culture and audience, by selling out culture.
(Adorno and Eisler 1994 [1947], 57)
It is worth reiterating, however, that the first screening of the Rosenkavalier

film took place in a prestigious opera house, not a picture palace, and that (as
we saw in Chapter 1) more elaborate presentations of silent feature films were
often mounted in similar venues. The boundaries between art and popular
entertainment were becoming blurred even at this early stage, and Adorno’s
and Eisler’s ‘breach’ between the allegedly self-contained audiences for both
is clearly an over-simplification.
Various attempts were made in the 1920s to improve synchronization
between sound and image in operatic films, notably in Germany, and these
included Blum’s rhythmonome (see Chapter 2). In 1922 the first opera
specifically intended for the silver screen, Ferdinand Hummel’s Jenseits des
Stromes, included musical notation as part of the projected image as a guide
to the conductor, but the film has not survived (Evidon 1992, 196; Fawkes
2000, 27). At the Vitaphone launch in New York in 1926 the feature film
Don Juan was prefaced by a series of short films of musical and vaudeville
performances: several star singers from the Metropolitan Opera had been
signed up by Vitaphone specifically to make synchronized shorts of popu-
lar operatic excerpts. As sound-on-film technology developed rapidly after
1927, Hollywood’s penchant for musical comedy resulted in the produc-
tion of many filmed operettas in the USA, while films of ‘serious’ operas
appeared mostly in Europe. An early highlight, which trod a middle ground
between the popular and the esoteric, was G. W. Pabst’s interpretation of
Weill’s The Threepenny Opera, shot in 1930 in both German and French ver-
sions using two different casts (Hinton 1990, 42–3). Pabst was a pioneer of
cinematic neue Sachlichkeit, his innovative montage techniques developing
methods of continuity editing still prevalent today. As with Max Ophüls’s
film of Smetana’s Bartered Bride (1932), Pabst showed how a front-rank
director could significantly enhance the impact of a stage work by skilfully
adapting it to the new medium. The script for The Threepenny Opera was
partly the work of Balázs, librettist of Bartók’s Duke Bluebeard’s Castle in
addition to being a noted film theorist, and the cinematography revelled in
the restless, searching camera movements, high-contrast lighting and shady
settings typical of Weimar cinema. Although the musical content was dras-
tically pruned, it was treated inventively throughout: creative use was made
of diegetic cues, as when ‘Mack the Knife’ is accompanied by a barrel organ
4.1 G. W. Pabst’s film of Kurt Weill’s and Bertolt Brecht’s The Threepenny Opera (1931) stimulated
sheet-music sales, the work’s publisher (Universal Edition) issuing this tie-in edition of four of the
film’s songs.
in a street scene (with the recording level manipulated to suggest distance in

a long shot), and songs used in instrumental versions played in a tavern and
brothel. A débâcle surrounding the film’s contractual arrangements drew
attention to the ongoing dangers in tackling copyright material: both Brecht
and Weill were legally entitled to have exclusive control over alterations to
the screenplay and music respectively, and both independently took the
production company to court when their entitlement was openly flouted.
Brecht lost his case, but Weill won his – securing in the process a hefty
cash settlement and potentially lucrative options to score further films by

the same company. On the film’s release in 1931, Universal Edition issued
a tie-in album containing four of the score’s most popular songs (Hinton
1990, 44–6).
In the UK, the 1930s saw the production of two expensive opera films in
colour, one based on Leoncavallo’s Pagliacci (destroyed by the distributor,
Trafalgar Films, after its completion in 1937 owing to the lack of revenue
generated by its release) and the other of Gilbert’s and Sullivan’s The Mikado
(dir. Victor Schertzinger, 1939). The latter was intended as the launch vehicle
for an ambitious series of G&S films featuring the D’Oyly Carte company
and London Symphony Orchestra, but the series was halted by the outbreak
of the Second World War. The Mikado met with mixed reviews, one critic
declaring that ‘the mechanical nature of the screen-photograph (added to
its self-complete realism in its own sphere) precludes any direct inter-action
between audience and performers’ (quoted in Huntley [1947], 48). The most
celebrated post-war British opera film was Michael Powell’s and Emeric
Pressburger’s interpretation of Offenbach’s The Tales of Hoffmann (1951),
in which dancing and powerful special effects combined to create a new
kind of cinematographic theatricality that appeared to be located in a fruitful
middle ground between fantasy and reality, described by film theorist André
Bazin as ‘an entirely faked universe . . . a sort of stage without wings where
everything is possible’ (quoted in Joe and Theresa 2002, 51). The film was a
major influence on director Martin Scorsese, its techniques directly affecting
his production methods in films as varied as Raging Bull, Taxi Driver and
New York, New York (Thompson and Christie 1989, 6). Also in 1951, a film
of Menotti’s The Medium received critical acclaim and subsequently won
an award at the 1952 Cannes Film Festival; this was an appropriate accolade
for Italian filmed opera in general, the country having remained at the
forefront of cinematic treatments of grand opera throughout the previous
two decades. Like Weill’s, Menotti’s musico-dramatic instincts were located
somewhere between the popular and the sophisticated, and for a brief time
in 1947 he had found himself under contract to MGM during the heyday
of that studio’s production of musicals, but left when his script entitled The
Happy Ending was not adopted – this scarcely coming as a surprise since the
happy ending concerned involved a group of children leaving their horrible
grandmother to freeze outside at Christmas (Fordin 1996, 222). A few years
later Menotti would strike a more universal chord with the success of his
warm-hearted Yuletide opera, Amahl and the Night Visitors (NBC, 1951),
the first opera written specifically for television.
Debate had in the 1930s begun to rage on the apparently fundamen-
tal tension between filmic realism and stage theatricality which concerned
Balázs. Because of the conceived incompatibility of the two approaches,
several early commentators on film music bluntly predicted no future for

filmed opera. Sabaneev identified the principal stumbling block as ‘the fact
that the art of the cinema . . . is a photographic art, and is therefore obliged
to be naturalistic and anti-theatrical’ (Sabaneev 1935, 26). London declared
filmed opera to be ‘impossible and intolerable’, and continued:
Those elements for which on the operatic stage even to-day allowance is
made, under the influence of the personalities of live artists, must on the
screen have an insipid, ridiculous, and anachronistic effect. The camera
brings the singer’s pathos much too close to the spectator; a close-up of a
photographed high C, on which the distorted face of the tenor, with
wide-open mouth, is to be seen, at once destroys the effect of even the most
beautiful melody and resolves it into laughter or even disgust.
(London 1936, 139–40)
In his account of the problematic nature of filmed opera, London concluded:

‘The unreal world of opera and the naturalistic film have nothing whatever
in common’. This view was elaborated by Kracauer in the early 1950s, when
he declared ‘The world of opera is built upon premises which radically defy
those of the cinematic approach . . . Opera on the screen is a collision of two
worlds detrimental to either’ (quoted in Joe and Theresa 2002, ix). London’s
assertions that ‘opera is static, film dynamic’ and that ‘well-known works of
operatic literature have become rigid conceptions which may not be touched
by the film’ (1936, 140, 142) will amuse contemporary opera audiences,
accustomed as they are not only to a spectacular diversity of production
styles beyond the wildest dreams of both film and opera audiences in the
1930s but also to the expectation that opera singers must be able to act
convincingly as well as to stand still and sing.
As early as 1913, Schoenberg had explored the possibility of filmed opera
in an unnaturalistic style when contemplating a hand-tinted silent film of
Die glückliche Hand. Writing to his publisher, he characteristically stipu-
lated that he should retain total control over all aspects of the live musical
performance, and showed himself to be in sympathy with the exigencies
of movie distribution by being prepared to consider the use of a cinema
organ instead of an orchestra if dictated by the size of the projection venue.
Schoenberg’s comments on the style of the visual images were far-sighted
in their experimental nature and awareness of the unlimited potential for
illusion inherent in the medium of film:
the basic unreality of the events, which is inherent in the words, is

something that they should be able to bring out even better in the filming
(nasty idea that it is!). For me this is one of the main reasons for considering
it. For instance, in the film, if the goblet suddenly vanishes as if it had never
been there, just as if it had simply been forgotten, that is quite different
from the way it is on the stage, where it has to be removed by some device.
And there are a thousand things besides that be easily done in this medium,
whereas the stage’s resources are very limited.
My foremost wish is therefore for something the opposite of what the
cinema generally aspires to. I want:
The utmost unreality! (Hahl-Koch 1984, 100)
Significantly, Schoenberg was considering Roller as one of three possible

designers (he was in good company: the other options were the expressionist
painters Kokoschka and Kandinsky), having been impressed by the anti-
realist tendencies he had shown in his production of Tristan for Mahler at
the Vienna Opera in 1903. Roller’s positive attitude towards cinema early in
the century was shown by a striking statement he made six years later in an
essay bemoaning what he perceived as a general lack of interest in theatrical
experimentation on the part of opera directors: ‘So why do we stick to a kind
of theatrical activity which seems no longer to be viable? Granted, new and
contemporary forms are continually arising, but in their lack of tradition
they are naturally not exalted enough to meet with serious encouragement
or to win favour with the cultivated! Isn’t a good film to be preferred to a
bad performance of Schiller?’ (Roller 1909).
A prolific output of filmed opera was produced by the Soviet Union from
the 1950s onwards, in tandem with a series of films of well-known Russian
ballets. Following the stage-bound film versions of Rachmaninov’s Aleko
(dir. N. Sidelev, 1953) and Mussorgsky’s Boris Godunov (dir. Vera Stroyeva,
1955), more creative cinematography was demonstrated in Tchaikovsky’s
Eugene Onegin (dir. Roman Tikhomirov, 1958). Tikhomirov later directed
films of Tchaikovsky’s Queen of Spades (1960) and Borodin’s Prince Igor
(1971). Other treatments of standard repertoire items included two films
by Vladimir Gorikker, of Tchaikovsky’s Iolanta (1963) and Rimsky’s Tsar’s
Bride (1964). The most notable Soviet opera film was based on a twentieth-
century score: Shostakovich’s Katerina Izmailova (the revised version of
Lady Macbeth of Mtsensk, withdrawn in 1936 after the composer’s infamous
lambasting in Pravda). The opera was filmed in 1966 by director Mikhael
Shapiro, working in close collaboration with the composer, and Galina Vish-
nevskaya took the title role as she had in the revised opera’s first staging in
1963. Shapiro combined sophisticated montage techniques with two fea-
tures common in later Soviet opera films: realistic settings and a double
cast (one of actors, the other of dubbed singers; the only exception was
Vishnevskaya, who fulfilled both functions). According to Tatiana Egorova,
the film’s only ‘serious mistake’ was the exclusion of naturalistic sound and
sound effects from the final soundtrack, with the result that ‘the visual ele-
ment of the film resembled an animated illustration, a pantomime of the
recorded opera’ (Egorova 1997, 190).
Outside the Soviet Union, the output of filmed opera had dwindled
somewhat, being largely confined to straightforward filmings of staged pro-
ductions, such as Paul Czinner’s films of the Salzburg productions of Don
Giovanni (1955) and Der Rosenkavalier (1961), or films designed for the
greater intimacy afforded by the medium of television, such as Ingmar
Bergman’s Die Zauberflöte (1975) and Jean-Pierre Ponnelle’s Le nozze di
Figaro (1976). Also dating from 1976 was a more ambitious project in which
Jean-Marie Straub and Danièle Huillet directed an austere outdoor version
of Schoenberg’s Moses und Aron that ‘represented the zenith of the Brechtian
anti-aesthetic trend in cinema’ (Joe and Theresa 2002, 215). This film was
too esoteric in both its choice of opera and style of presentation to be widely
influential, and interest instead began to focus on the commercial viability
of straightforward treatments of popular operas in lavish period settings.
The 1980s vogue for full-scale grand opera in the cinema was initiated
by Joseph Losey’s film of Don Giovanni (1979), performed on the sound-
track by the Paris Opéra under Lorin Maazel. The project was conceived by
Rolf Liebermann, who considered Patrice Chéreau and Franco Zeffirelli as
possible directors before deciding on Losey – who had never seen the opera.
The action was shot entirely on location amongst the impressive Palladian
architecture of Vicenza and in the Venetian islands, and viewers who merely
revelled in the visual splendour of the cinematography may have been sur-
prised to learn that the stunning locations were used ‘to erect a Marxist
critique of class relations’ (Citron 2000, 11–12). Critical responses ranged
from Julian Rushton’s curt dismissal of the project as an ‘elegant imbecility’
(Rushton 1981, 80) – a remark which might also be applied to a good deal
of the operatic repertoire itself, even in its unfilmed state – to David Caute’s
attempt to prove that this ‘masterpiece ravishing to both ear and eye’ is
dramatically superior to a stage interpretation:
Losey met the challenge by flooding the picture with sunlight and water,
paintings and sculpture, his camera movements boldly responsive to
Mozart’s music, a dazzling fusion of the fine and performing arts.
Confronted by the visual stasis of operatic convention, Losey eased apart the
orchestral and the dramatic, reuniting them in the cutting-room on his own
terms. (Caute 1994, 431)
At the time of the film’s release, the critical response in the UK and USA
was almost unremittingly negative. Only in France did massive publicity on
the part of the producers (Gaumont) help the film to score an enormous
success at the box office and receive the critical adulation that eluded it
elsewhere. This was in spite of a major rift between Liebermann and Losey
when the former objected to the latter’s inattention to matters of dynamics
in the score, and his tendency (in marked contrast to Shapiro’s in Katerina
Izmailova) to allow sound effects to dominate the singing; as Liebermann
put it, unconsciously echoing Breil’s defence of Wagner, ‘Mozart did not
compose film music’ (Caute 1994, 430–1).
Zeffirelli brought his considerable stage experience, as both director and
designer, to bear in a film version of Verdi’s La traviata (1982), with a
music track performed by the Metropolitan Opera under James Levine.
Characteristic of its director were the opulent costumes and sumptuous
interiors with warm lighting, and the simple use of stock cinematic devices
such as flashbacks (to illustrate a character’s thoughts), slow zooms in and
out, and the occasional use of voice-over in soliloquies – the last a neat way
of avoiding the intrusiveness of close-up photography of singing mouths. In
the opening sequence, a flash forwards to the dying Violetta’s dustsheet-clad
apartment, and again during the Prelude to Act III, the combined effect of
the mute visual images and Verdi’s heart-on-sleeve music was remarkably
similar to that of silent-film melodrama of the 1910s. A follow-up film
of Verdi’s Otello (1986), conducted by Maazel, took significant liberties
with the score in the interests of serving the visual images, to the extent
that the defensive director prevented music critics from attending the New
York première (Citron 2000, 74). Otello failed to repeat the success of La
traviata, even though (like the same director’s Romeo and Juliet of 1968)
it predictably netted the Academy Award for best costumes. In a television
interview in 1997, Zeffirelli claimed that he felt opera to be the most complete
artistic form, combining ‘dance, drama, poetry, music and the visual arts’
(R. Jackson 2000, 212); yet, in spite of their surface gloss, his filmed operas
were deeply conservative in their production values and offered little to
stimulate either the intellect or the emotion, treading a careful middle-
ground between restraint and excess.
More successful was Francesco Rosi’s version of Bizet’s Carmen (1984),
which also used traditional costumes and appropriate exterior locations.
Produced by Gaumont and again featuring a music track conducted by
Maazel, this flamboyant and colourful interpretation was distinguished by
effective crowd scenes, several of which used voice-overs instead of mimed
singing to achieve greater visual realism and freedom of movement. The
main titles appear over a slow-motion bullfight to the accompaniment of
crowd noise and distant snatches of music; the overture crashes into life at
the precise moment when Escamillo’s sword enters the bull’s neck and the
animal drops lifeless to the ground. An effective sequence in its own right,
this prologue forms a symmetrical counterpart to the bull-fighting climax
with which the film concludes. Bizet’s score proved to be admirable for
cinematic adaptation in those instances where purely instrumental passages
could be treated by Rosi as straight underscoring to the action on screen. As
H. Marshall Leicester has shown, Rosi’s refined cinematography – which at
times purports to be as realistic as Zeffirelli’s – subtly underlines the contrast
in the score between vernacular musical idioms and conventional operatic

gestures: ‘Rosi converts “realism” into a textual element, making use of the
rich reference to actuality in a way Zeffirelli, the exemplar he emulates and
parodies, never dreamed of. He exploits the oddness of people singing in
what looks so much like real life to specify and reinforce the psycho-social
implications of a difference in musical style’ (Leicester 1994, 273).
The diversification of opera films in the 1980s resulted in several con-
trasting ventures entirely different from the heady energy of Rosi’s Carmen.
Hans Jürgen Syberberg’s interpretation of Wagner’s Parsifal (1982), played
out on a massive set modelled on the composer’s death mask, combined
boredom and pretension in equal measure; singer Robert Lloyd (who played
Gurnemanz) recalled that ‘none of the cast really understood what Syber-
berg was on about, they simply did what he asked’ (Fawkes 2000, 182). The
portmanteau film Aria (1988) was a hotchpotch compilation of operatic
excerpts interpreted by no fewer than ten different directors with varying
degrees of flair and success; according to Derek Jarman, ‘not one of the direc-
tors opened the music up; in nearly every case they used it as a backdrop for
a series of rather arbitrary fantasies, none of which had the depth or com-
plexity of the original work’ (Jarman 1989, xi). (For a detailed audio-visual
analysis of the segment based on Lully’s Armide, directed with characteristic
idiosyncrasy by French New Wave director Jean-Luc Godard, see N. Cook
1998, 215–60.) Losey had been planning a film of Tosca before his death
in 1984; Puccini’s opera received an over-ambitious live-on-location film-
ing in 1992, and an accomplished interpretation by director Benoı̂t Jacquot
in 2001 in which full-colour costumed scenes in luminous Zeffirellian set-
tings were disconcertingly intercut with monochrome footage of the singers
recording the soundtrack in modern dress – a rather gratuitous reminder
that the singing in the film is, as usual, entirely pre-recorded.
Film in opera; opera in film

Berg’s Lulu (1935) was one of several stage operas that imported a filmic
element into their dramaturgy. Based on plays by Wedekind filmed by Pabst
in 1929, Lulu features at its mid-point a palindromically scored film interlude
depicting Lulu’s incarceration and subsequent release from jail: this device
serves both as a convenient compression of stage time and as a graphic
illustration of the turning point in the drama before various elements in
the second half start to run in reverse. Filmed segments also featured in
Satie’s ballet Relâche (1924), in Hindemith’s Hin und zurück: eine Zeitoper
(1927) and in a production of Wagner’s Ring in Berlin (1928). In Milhaud’s
Christophe Colomb (1930), according to the composer’s wife, ‘for the first
time in an opera one could see moving films showing scenes a little different
from what was happening simultaneously on stage’ (Nichols 1996, 39).
In 1994, Glass conceived his La Belle et la bête as a simultaneous (silent)
projection of Cocteau’s 1946 (sound) film of the same title with a new
musical accompaniment provided by live but static singers (Joe and Theresa
2002, 59–73). Glass’s novel venture was a follow-up to his Orfée (1993), an
operatic version of another film by Cocteau, and caused considerable critical
controversy in France (Walsh 1996).
Conversely, a steady succession of narrative films featured operatic
excerpts in their screenplays. An opera-house setting was memorably used
in the silent film of The Phantom of the Opera (dir. Rupert Julian, 1925),
starring Lon Chaney, and remade with sound in 1943. The Marx Brothers
incorporated an extended segment of Il trovatore in their riotous comedy A
Night at the Opera (1935), but later cinematic appropriations of opera were
generally serious in intent. The diegetic opera in Citizen Kane (see Chapter
5) was a rare example of specially composed operatic music: more typi-
cal has been the use of pre-existing operas from the popular repertoire. In
the film version of Peter Shaffer’s play Amadeus (dir. Milos Forman, 1984),
numerous lavishly staged extracts from Mozart’s operas provided spectac-
ular punctuation to the drama but somewhat impaired the narrative flow.
Elmer Bernstein’s score to Scorsese’s The Age of Innocence (1993) evoca-
tively superimposed haunting original music onto a diegetic performance
of Gounod’s Faust in order to suggest the romantic enchantment of the two
protagonists as they meet in a box at the theatre. In The Godfather Part III
(dir. Coppola, 1990) a diegetic performance of Mascagni’s Cavalleria rusti-
cana forms an ironic and cohesive backdrop to a climactic series of killings.
Nino Rota’s music for the two previous films in the Godfather trilogy (1972
and 1974) drew heavily on the bel canto idiom of Italian opera, and Verdi’s
music in particular remained a popular choice for film-makers outside a
specifically Italianate setting (see Chapter 9).
Several idiosyncratic uses of operatic subject-matter are to be found in
non-Anglophone films. Krzysztof Kieślowski’s Personnel (1975), in which a
young man for whom opera represents a pinnacle of fantasy has his illu-
sions cruelly shattered by the mundanity of working behind the scenes at
an opera house, plays out a political metaphor of repression in Poland:
according to the director, ‘our dreams and ideas about some ideal reality
always clash somewhere along the line with something that’s incomparably
shallower and more wretched’ (Stok 1993, 96). In Jean-Jacques Beineix’s
Diva (1981), a fictional opera star is idolized by a young lad who records her
singing illicitly (see Chapter 8), and much play is made on the artificiality
of nondiegetic music in film. (Diegetic and nondiegetic music in film are
broadly comparable to Carolyn Abbate’s conception of the ‘phenomenal’
and ‘noumenal’ in opera: see Abbate 1991.) In Werner Herzog’s Fitzcarraldo

(1982), the building of an opera house in the middle of a Peruvian jungle
is ‘paradigmatic of the desire for an opera and for an art generally to be
situated outside the commercial pressures that have previously constituted
the arts’ (Tambling 1987, 18; see also Rogers 2004). Federico Fellini’s E la
nave va (And The Ship Sails On, 1982) is a delicious parody of operatic
decadence on board an ocean liner heading out to sea to scatter the ashes of
a deceased diva in 1914: the music of Verdi is poised fantastically between
the diegetic and the nondiegetic in the manner of a film musical, and the
production celebrates the operatic medium’s inherent theatricality not only
in the two-dimensionality of its principal characters but by using patently
flimsy cardboard cutouts and crude models of ships set in a heaving sea
of plastic sheeting. The not inconsiderable poetry created by the sumptu-
ous cinematography and music – artificial sets and stereotypical characters
notwithstanding – is utterly negated at the close by the camera’s eye turn-
ing (in typical New Wave fashion) away from the shipboard set to reveal
the mechanics of the modern film studio in which the illusion has been
mounted. Particularly iconoclastic was Alexander Kluge’s deconstruction
of both operatic music and its institutionalization in The Power of Emotion
(1983), in which the imposing architecture of the opera house, like both fac-
tories and power stations, itself represents a bastion of authority generating
dubious emotional outpourings (Flinn 2004, 151).
Disembodied recordings of operatic performances also figure promi-
nently in narrative films. In The Shawshank Redemption (dir. Frank
Darabont, 1994), a recording of Mozart’s Figaro is relayed through prison
loudspeakers to symbolize freedom – and this universal message is instantly
comprehended by all classes of inmate. An identical device appears in Life
is Beautiful (dir. Roberto Benigni, 1997), where recorded Offenbach sim-
ilarly lifts the spirits of those incarcerated in a Nazi concentration camp.
A recording of an aria from Giordano’s Andrea Chénier illuminates vari-
ous aspects of character and cultural context in Philadelphia (dir. Jonathan
Demme, 1993). All these instances use an outdated musical idiom to suggest
a sense of timelessness or nostalgic yearning, much in the fashion of the aural
utopia conjured up by old-fashioned Hollywood film music. But, as Marc
A. Wiener points out, operatic music can be used in films for diametrically
opposed purposes according to context: ‘When it represents particularity,
opera signifies entrapment, and when it functions as a sign of the universal,
it represents freedom’ (Joe and Theresa 2002, 83).
The ongoing links between cinema and opera have been fostered not
only by composers working in both genres, but also by a clutch of influential
directors and designers whose work straddles both media. Directors of both
staged opera and feature films include Baz Luhrmann, Chéreau, Eisenstein,
Herzog, Nicholas Hytner, Losey, Rouben Mamoulian, Ken Russell, Luchino

Visconti and Zeffirelli. Film designer Georges Wakhevich worked on Peter
Brook’s 1948 Covent Garden production of Mussorgsky’s Boris Godunov – a
staging which Ernest Newman explicitly condemned for its filmic qualities
(Sutcliffe 1996, 20–1). Some opera singers have made a significant impact as
screen actors, including Maria Callas in a magnificent interpretation of the
title role in Medea (dir. Pier Paolo Pasolini, 1970) – described by one critic
as ‘an opera without music’ (J. Walker 2006, 750) – and Teresa Stratas in a
powerful performance as a self-absorbed singer who neglects her autistic-
savant daughter in the Canadian production Under the Piano (dir. Stefan
Scaini, 1995). Stratas had previously received critical acclaim for her inter-
pretation of the title role in Götz Friedrich’s film of Strauss’s Salome (1974).
The first and virtually the only opera specifically conceived as a sound
film was The Robber Symphony by Friedrich Feher (1936), and in those
heady days of the early sound film it seemed to bode well for the genre’s
future. An anonymous contemporaneous critic commented: ‘The intimate
alliance of music and fantasy is in principle wholly commendable . . . Here,
perhaps, may even be opera’s legitimate successor – the transmutation of that
hitherto over-synthetic medium into something more complex and closely-
knit’ (quoted in Huntley [1947], 45). Ernst Toch went further, predicting
that
The focus of film music to come is the original film opera. This cannot be
done by adapting old operas for the screen, for the conception of
stage-opera music is bound to be different from what film-opera must be.
To adapt existing operas . . . means to mutilate either screen action or the
music itself. Music of film-opera has to create and develop its own forms
out of typical screen action, combining its different laws of space, time and
motion with constant music laws. The first film-opera, once written and
produced, will evoke a host of others. (Toch 1937)
Two years after these lines were written, Shostakovich was ‘contemplating
writing a film-opera, exploiting the principles of realism to the full’, con-
fessing himself to be ‘very attracted to the limitless possibilities opened up
by the cinema screen’ and citing Prokofiev’s Alexander Nevsky music as an
inspiration (see Chapter 9), but lamenting that all his attempts to secure
the necessary collaborators had come to naught (Shostakovich 1981, 78–
9). As we have seen, later filmed operas were invariably based on existing
repertoire, and the kind of creative vision demonstrated by Schoenberg’s
unachieved plans for Die glückliche Hand was reflected only in the Powell–
Pressburger Tales of Hoffmann in 1951. That same year marked the birth
of opera conceived for television, which bypassed some of the perceived
problems inherent in filming opera for the big screen, while presenting
new challenges of its own. Since the brief burgeoning of interest in tele-
vision opera in the 1950s and 1960s, when the dramatic style was steeped
in multi-camera studio techniques, more recent productions aimed at the
small screen have returned to an on-location realism and widescreen glossi-
ness influenced by cinematic techniques and shot on film rather than video.
The striking difference between the two may be examined by a comparison
between the BBC’s and Margaret Williams’s productions of Britten’s Owen
Wingrave, made in 1971 and 2001 respectively (see Cooke 2005, 285–9). As
recently as 2003 a television film of John Adams’s The Death of Klinghoffer
(dir. Penny Woolcock) for the most part cultivated a detailed cinematic real-
ism consistently at odds with the extreme stylization in the music. Adams’s
score cries out for a corresponding degree of visual stylization, and the
unnecessary realism in this film version paradoxically makes suspension of
disbelief well-nigh impossible. Realistic terrorists with realistic guns on a
realistic ship in a realistic ocean seem a viable cinematic proposition right
up to the moment when the characters open their mouths and break into
apparently diegetic song, and this problem seems far more acute in works
with a topical modern setting than in the case of, say, Losey’s period-dress
Don Giovanni where the stylization is a natural extension of setting and
ambience. The difficulties of negotiating the conflicting demands of cin-
ematic realism and operatic theatricality remain as strangely problematic
today as they did at the birth of the sound film.
The film musical

Operas specially written for film have always been exceptionally rare and,
although rather more have been commissioned for television, the combined
total is still small. This select repertoire has been notable for its surprisingly
low level of dramatic and visual experimentation – possibly a consequence
of understandable attempts on the part of those composers involved to write
works which would transfer relatively easily to the live operatic stage and
thereby stand a chance of securing a living future in the theatre. Clearly the
commercial non-viability of filming modern opera for exhibition in cinemas
has been a major factor contributing towards the paucity of original film
operas. But the roots of the problem of filming opera go right back to the
1930s when the film medium was generally perceived as too lowbrow to be
applied to the interpretation of great works of art. The debate that raged in
the 1940s on the subject of the relative merits of realism and theatricality
in filmed opera productions singularly ignored the fact that the stunningly
entertaining Hollywood musicals of the 1930s had already shown what could
be achieved when music and image were creatively combined and fantasy
allowed free rein, and it is a considerable irony that this degree of artistic
experimentation seemed acceptable only in the context of a popular and
commercially viable genre.
As we saw in Chapter 2, the success of the Jolson vehicles The Jazz Singer
and The Singing Fool derived from a combination of two of the least perma-
nent attributes of popular entertainment: technical novelty and star qual-
ity. Both films were to some extent rooted in silent-film techniques, and
inspired an immediate glut of similarly plotted ‘mammy’ pictures starring
Jolson, Eddie Dowling and French crooner Maurice Chevalier – the last
negotiating his transfer from the Parisian stage to Hollywood celluloid with
ease. Next came a glut of backstage scenarios which began to demonstrate
the extraordinary potential of the sound film for the production of more
flamboyant musical dramas, while to a large extent continuing the tradi-
tions of vaudeville and the titillating Gallic dance spectacles typified by
the annual Ziegfeld Follies (1907–31). The backstage genre not only rep-
resented a departure from the standard fare of the silents: it thrived on
a combination of spectacle, music, dance, romance, star attractions and
sheer virtuosity that put all rival media in the shade. A perceived need for
both songs and dance to be justified by the diegesis and the initial neces-
sity of recording music and sound live on set were two reasons why the
revue format was initially popular, and this blossomed into a widespread
craze for comic backstage plots which served to justify the inclusion of stage
routines. Critic Alexander Bakshy discussed what he viewed as an essential
incompatibility between theatrical comedy and these screen antics, opin-
ing that on stage a musical ‘disregards the absurdities of the plot or the
antics of the characters, because it never associates them with real life’ and
lamenting the fact that makers of film equivalents ‘place their characters
in perfectly natural surroundings, and introduce them as perfectly normal
people, and then make them behave as if they were escaped lunatics’ (Bakshy
1930a). He also noted that the mass-production of backstage films was not
driven by public desire but because ‘Hollywood finds in stage life the easi-
est formula for making a song and dance show realistically plausible’, and
went on to demand more cinematic imagination instead of merely ‘aping
the stage’ (Bakshy 1930b). For all their surface sophistication and candid
voyeurism, the early backstage releases ‘were simply glossy repackagings of
the Protestant work ethic, preachments of virtue and talent rewarded, and
proved to be the easiest, most elementary foundation for musicals’ (Bar-
rios 1995, 190). They nevertheless soon cross-fertilized with other genres,
notably the gangster and underworld movies for which Warners became
notorious.
The prototypical backstage musical was The Broadway Melody (dir. Harry
Beaumont, 1929), made by MGM and featuring songs with music by Nacio
Herb Brown and lyrics by Arthur Freed, dressed up with jazzy orchestration
akin to the style of popular bandleaders Paul Whiteman and Jean Goldkette.
Originally conceived in a part-talking format like the Jolson films but trans-
formed into the very first ‘all talking, all singing, all dancing’ spectacular
during production, the film won an Academy Award and (thanks to the
use of subtitles in export prints) fame abroad, netting some $4 million in
box-office takings. Shot mostly with four simultaneously running cameras
to facilitate soundtrack editing, the film departed from this procedure in its
Technicolor sequence, ‘The Wedding of the Painted Doll’, which embodied
a retake in which the music had not been re-recorded live: the Movietone
soundtrack from the first take had instead been played back on set and then
postsynchronized to the new image track. This major breakthrough in edit-
ing techniques had arisen from the studio’s desire to cut costs by not recalling
the musicians for the new take (Barrios 1995, 65). Although the camera work
at times offered a far more intimate experience to the spectator than would
have been possible in live theatre, The Broadway Melody nevertheless suf-
fered from a static and essentially theatrical approach to stage blocking and a
proscenium-like framing of its images. The film was at its creative best when
a diegetic pretext in the opening scene generated an exhilarating mélange of
Dixieland, stride and popular-song styles, all rehearsed simultaneously by
different performers in the adjoining offices of a music-publishing company.
Rehearsal scenes were to remain a staple ingredient of backstage musicals
precisely because they permitted this kind of informality, and served as a
useful halfway point between everyday realism and performative stylization.
Later, dream sequences became a stock method for introducing elements
more fantastic than diegetic realism might otherwise have permitted, inten-
sifying the feeling that the act of performance – whether realistically stage-
bound or surreal – constitutes a cathartic release from the humdrum cares
and responsibilities of both the cast’s and audience’s everyday lives (Altman
1987, 61).
The Broadway Melody spawned a series of sequels in the 1930s and a
direct remake with new songs (Two Girls on Broadway, 1940), plus a bizarre
MGM short called The Dogway Melody (1930) which satirized the studio’s
own product in the shape of a miniature backstage musical performed
entirely by anthropomorphic dogs, including a parody of ‘Mammy’ sung by
a black dog named Al J. Olsen and a fully dogeographed finale on Freed’s
and Brown’s song ‘Singin’ in the Rain’ – a recent hit from MGM’s The
Hollywood Revue of 1929. Direct competition to MGM immediately came
from RKO (Syncopation, dir. Bert Glennon), Paramount (Close Harmony,
dir. John Cromwell and Edward Sutherland), Universal (Broadway, dir. Paul
Fejos) and Warner (the all-colour On With the Show!, dir. Alan Crosland), all
released in 1929. Fox’s Sunny Side Up (dir. David Butler, 1929) was significant
not only for its original screenplay but also for its attempt to integrate musical
numbers into the narrative. Universal exploited the enormous popularity
of the Whiteman band and showcased it in the big-budget King of Jazz (dir.
John Murray Anderson, 1930), a lavishly entertaining film much vilified by
jazz scholars on racial grounds but including bold innovations such as an
animated sequence in colour.
Relative newcomer RKO typified the synergistic marketing of popular
music enabled by its commercial interests in vaudeville and particularly
radio, using the latter not only to plug its songs relentlessly but also to
broadcast complete performances of the Syncopation soundtrack, recorded
using the Photophone technology of its parent company RCA (Barrios 1995,
86–7). The major studios had significant commercial interests in popular
music publishing, their acquisition or foundation of sheet-music businesses
both freeing them from the need to pay punitive synchronization rights to
others and opening up opportunities for generating additional revenue.
The history of film musicals was thus indissolubly tied to a desire on the
studios’ part to minimize expenditure on musical copyright and maximize
dissemination of their own hit songs in print, airwave and recorded forms.
The revenue-generating power of the last had been spectacularly shown by
million-mark sales of Jolson’s recording of ‘Sonny Boy’, from The Singing
Fool. Towards the end of 1929 no fewer than 90 per cent of the most popular
songs in the USA were directly related to films. By 1939 approximately
two-thirds of the royalty payments made by the American Society of
Composers, Authors and Publishers (ASCAP) to publishing houses went
to those with Hollywood affiliations, and the earning power of a film musi-
cal might typically expect to be augmented by around $1 million thanks to
radio broadcasts, sheet music and record sales (J. Smith 1998, 31). When film
musicals were based directly on stage shows, the original hit songs might be
replaced by inferior songs produced in-house in the interests of economy,
and more durable studio songs could be recycled in later productions by
the studio which owned the rights (Barrios 1995, 108–11). A studio might
also acquire the entire song catalogue of a popular tunesmith, showcasing
the material in biopic musicals loosely based on the composer’s life story
or a fictionalized equivalent. Wilful distortion of the biographies of these
famous musicians in many films in this category (for jazz examples, see
Chapter 5) served to deflect attention from the commercial basis of much
of their creative activities. As Rick Altman puts it, ‘biographical events are
ignored in order to make the semantic givens of the biopic conform to the
syntax of the show musical. Music must never be seen as something one
does solely to make a living. To make music is to make love; to make love is
to inspire art’ (Altman 1987, 238).
4.2 Gold Diggers of 1933: Busby Berkeley’s chorines dance with electrified violins to ‘The Shadow Waltz’.
The backstage musical acquired stunning visual interest and choreo-

graphic flair at the hands of Busby Berkeley, who cut his teeth on Samuel
Goldwyn’s production of Whoopee! in 1930 and secured his reputation with a
series of spectacular dance routines for Warner Bros., including those in 42nd
Street (dir. Lloyd Bacon, 1933), Gold Diggers of 1933 (dir. Mervyn Le Roy),
Footlight Parade (dir. Bacon, 1933), Gold Diggers of 1935 (dir. Berkeley),
Gold Diggers of 1937 (dir. Bacon) and Gold Diggers in Paris (dir. Ray Enright,
1938). The first item of the Gold Diggers series was a remake of Gold Diggers
of Broadway (dir. Roy Del Ruth, 1929), a Vitaphone colour film that survives
incomplete, which had featured the hit song ‘Tiptoe Through the Tulips’
and netted Warners some $4 million. Warners’ ailing fortunes in the Depres-
sion years were comfortably revived by 42nd Street, with its sassy scripting,
superlative cast and impressive production values helping the project tie
directly in with the upbeat optimism of the new Roosevelt administration.
Virtuosic as much in their novel camerawork (involving elaborate tracking
shots and birdseye crane views) and stylish editing as for their infamously
fetishistic and elaborate chorine formations, Berkeley’s choreography and

visual flair in his interpretations of songs by Harry Warren and Al Dubin
opened the backstage genre out into the realm of pure fantasy – though
it is easy enough to dismiss his work for giving ‘ingenuity and vigor . . .
rampant precedence over taste’ (Barrios 1995, 247). Highlights include the
ubiquitous cash props in ‘We’re in The Money’, the multiple neon-glowing
electrified violins of ‘The Shadow Waltz’ in Gold Diggers of 1933, and the
massed ranks of no fewer than 52 swirling white grand pianos in ‘The Words
Are in my Heart’ sequence in Gold Diggers of 1935. Stephen Banfield has
noted how the ‘modulor’ principle in popular-song composition permitted
basic 32-bar AABA structures to be extended to considerable lengths in the
process of instrumental arrangement for the purposes of choreographing
set-pieces such as these (Banfield 1998, 333).
Arguably the finest of the earliest show films was Paramount’s Applause,
released in the boom year 1929, which embodied Mamoulian’s experimen-
tation with mobile cameras and novel sound combinations, and showed
a reluctance to buy slavishly into the genre. Mamoulian approached even
closer to Clair’s ideal of the integrated musical film in Love Me Tonight, a
Paramount vehicle for Jeanette MacDonald and Chevalier made in 1932,
but his imaginative ideas for the genre were later submerged. When he
subsequently worked for MGM and was asked in 1945 by Freed, now a
powerful producer, to direct a musical version of Eugene O’Neill’s play
Ah, Wilderness!, Mamoulian demonstrated his awareness of the essential
incompatibility between stage technique and cinematic drama: ‘If you take
something written for the stage and put it on the screen, you’re going to lose
certain values. Now, if you cannot compensate for these values, add to them,
make them expand and flourish – then you shouldn’t touch it’ (quoted in
Fordin 1996, 186). In a later memo he expanded on the theme:
It is obvious that to turn a dramatic play into a musical you have to make
drastic cuts in it in order to allow time for music, songs and dancing. It is
equally obvious that you cannot drastically cut a good dramatic play
without spoiling it, crippling its subject and emaciating its characters. A
tenuous story filled out with elaborate and overblown ‘musical numbers,’
unrelated specialties of dance and song, comedy routines, etc. has for long
been the standard stuff that musicals were made of . . . However, this is not
the kind of musical we want to make . . .
What we want to do is, for the lack of a better and newer definition, a
‘musical play’ – meaning by that a story which will be told through the
medium of integrated dialogue, songs, dance and music, with each of these
elements taking an organic and vital part in the telling of that story. What
happens in this case is that the dialogue scenes, which have been cut out of a
good play, are not thrown overboard, but are actually translated into their
musical equivalent of song and dance. As a result the story has not suffered,
nor has it changed, but the manner of telling it has changed, and it has
been enriched by added emotional values which the right kind of music
brings. (Fordin 1996, 188)
In this instance the resulting film (Summer Holiday, 1948; songs by Harry
Warren and Ralph Blane) was one of MGM’s few flops, losing $1.5 million
at the box office. Although this failure might ostensibly have suggested a
doomed future for creative experimentation, Mamoulian’s work had a sig-
nificant influence on the later development of musico-narrative integration
in the genre.
In his critical analysis of the American film musical (1987), Altman iden-
tifies two principal subgenres in addition to the show category (which he
extends from backstage examples to embrace musical films concerned with
mounting many different kinds of performance): the fairy-tale musical and
the folk musical. The fairy-tale subgenre, to which Mamoulian’s Love Me
Tonight belongs, was a natural development from popular operetta, with
its old-fashioned waltz-dominated style retained from its former European
stage incarnation, and first arrived on the silver screen in the shape of Warner
Bros.’ The Desert Song (dir. Roy Del Ruth, 1929; remade in 1943 and 1953).
This 1926 stage show by Sigmund Romberg and Oscar Hammerstein II
transferred to film with the help of a Vitaphone score recorded on set under
the musical direction of Louis Silvers; like rival products, the film included a
Technicolor sequence. (Many of the colour segments of early films have not
been preserved, and survive only in far more durable monochrome prints
prepared for archival purposes. When sold to television companies in the
1950s, only monochrome prints were useful for broadcasting and colour
prints were mindlessly destroyed: see Barrios 1995, 435 n.1.) Even in 1929,
reviewers lamented the stage-bound nature of The Desert Song transfer and
yearned for a more cinematic treatment. RKO’s contemporaneous hit trans-
fer Rio Rita (dir. Luther Reed) was just as wooden in its theatricality, but
at least featured some location shooting. Stilted films of fairy-tale operettas
seemed poised to occupy a uniquely middlebrow niche at a time when mod-
ern popular styles were still tainted by their sometime insalubrious vernac-
ular origins. The first written expressly for the screen was Paramount’s The
Love Parade (1929), starring Chevalier and MacDonald with music by Victor
Schertzinger, which benefited from the lightness of touch and deft wit of
expatriate German director Ernst Lubitsch, among whose innovations was
the thorough organization of actors’ movements according to regular rhyth-
mic beats in order to facilitate synchronization with the music. Warner Bros.
squandered its massive profits from the early sound boom on miscalculated
operetta projects when the industry’s Depression slump was imminent, the
economic disaster occurring at the same time as the sound film lost its ini-
tial novelty value. After the fallow years 1931–4, creative experimentation
dwindled and the musical’s subgenres were destined to become even more
rigidly defined and constantly regurgitated.
The Love Parade initiated a trend towards romantically pitting contrast-
ing voice- and character-types against one another in the kind of binary
opposition which, as Altman has argued (1987, 16–27), replaces linear nar-
rative structure in many later musicals. The ‘dual-focus’ narrative struc-
ture he identifies can variously involve contrasts between the two sexes,
two attitudes (e.g. work versus play, manifested specifically as business ver-
sus entertainment in Silk Stockings (1957), and seriousness versus fun in
both Funny Face (1956) and The Sound of Music (1965)), or two classes –
often shown by two different singing styles, an operatic aristocratic female
and a rough down-to-earth male, or by two dancing styles, as with the
contrast between Fred Astaire’s tap-dancing and Ginger Rogers’ classical
elegance. Narratives of this kind achieve eventual union between the oppos-
ing poles not through the chronological and causal linear structures com-
mon in other genres, but through simultaneity and comparison, as in the
parallel scenes and songs for the male and female leads in MGM’s Gigi
(1958). The dual-focus approach explains why the apparently intrusive set-
piece routines are vital in developing and exploring the binary oppositions
rather than interrupting a (non-existent) linear narrative to its detriment.
Vocal duets and dancing couples not surprisingly become central to this
discourse.
The folk musical began its screen life with Universal’s patchwork pro-
duction of Show Boat, which started out as a silent drama and was rejigged
in 1929 with interpolated songs following the success of Jerome Kern’s and
Hammerstein’s stage musical; remakes followed in 1936 (dir. James Whale)
and 1951 (dir. George Sidney). The Kern–Hammerstein treatment was of
incalculable importance in demonstrating how musical numbers could be
justified by plot and character, and exerted a powerful influence on the
development of the so-called ‘integrated’ musical (in contrast to the cumu-
lative effect of the loosely episodic ‘aggregate’ show); their work also broke
new ground in departing from operetta stereotypes in search of a more
conversational musical style derived from vernacular idioms. According to
Banfield, the stage show was ‘more a foretaste of what the sound film might
be expected to attempt with popular music than the theatrical breakthrough
it is always taken to be . . . Its epic dimension . . . boils down musically to
a superior silent-film technique of motivic underscoring’ (Banfield 1998,
327). In this and other folk musicals, singing becomes natural expression
rather than stylized performance, and is more often than not rooted in the
everyday activities of an enclosed community. The new style was evident in

Mamoulian’s treatment of Kern’s and Hammerstein’s songs in his own ven-
ture onto folk territory, High, Wide and Handsome (Paramount, 1937). With
the addition of a chorus of ordinary folk indulging in joyous dance routines,
the folk subgenre became indestructible, and it owed this element largely
to the 1943 Broadway production of Richard Rodgers’ and Hammerstein’s
Oklahoma!, which had been directed for the stage by Mamoulian.
A unique contribution to the early film musical was King Vidor’s Hal-
lelujah! (MGM, 1929; musical direction by Eva Jessye), a folk project con-
siderably ahead of its time in its creative semi-documentary portrayal of the
everyday lives of black workers. In the course of Vidor’s provocative and
thoughtful dramatic intertwining of sexual and religious fervour, musical
expression in a variety of ethnic genres (including spirituals, the blues and
jazz) likewise seemed a direct extension of both character and circumstance.
The protagonist Zekeil becomes a preacher and sermonizes in a melodious
Sprechstimme that can easily blossom into fully blown song, while com-
munity dance and song seem an entirely natural means of expression; the
intensity of the singing in the spiritual numbers and the almost minimalist
riffs of the diegetic banjo song accompanying dancing children are unforget-
table. MGM returned to the all-black musical with Cabin in the Sky (1943),
the first film directed unaided by Vincente Minnelli, soon to become the
artistic lynchpin of MGM’s musicals production unit headed by Freed. Pit-
ting rural simplicity against the corrupting power of urban sophistication,
Cabin in the Sky in the process showed how racial stereotyping had if any-
thing deteriorated since Vidor’s generally sympathetic film. Now the focus
of interest was on a kind of commercially successful jazz, represented in the
film by the music of Duke Ellington and others, an attitude reflecting what
James Naremore has described as ‘a chic, upscale “Africanism,” redolent
of café society, Broadway theater and the European avant-garde’ and which
Krin Gabbard summarizes as ‘uptown Negrophilia’ (Gabbard 1996, 181–2).
In the same year, Fox’s Stormy Weather (dir. Andrew Stone) revisited both
backstage and biopic formulae in the shape of an all-black revue featuring
such stellar talents as Fats Waller, Cab Calloway and Lena Horne.
Plots in later white folk musicals were founded in a nostalgic regard for
family values, usually in a small-town or rural setting, and tapped memories
of familiar period tunes and iconography from the graphic arts and other
cinematic genres (e.g. the western) as appropriate. The turning point was
MGM’s Meet Me in St Louis (dir. Minnelli, 1944), superbly inventive visually
and with an integrated score including both original songs by Martin and
Blane and arrangements of fondly remembered period tunes such as the
lilting title song and ‘Skip to My Lou’. The dynamic staging of the ‘Trolley
Song’, a complex routine perfected by Judy Garland in a single take, set a new
standard in rhythmicized vernacular song style directly inspired by physi-

cal motion. According to Flinn, it is dramatically significant because ‘where
non-representational signs work so strenuously to convey an advanced tech-
nological utopia, the representational component of the lyrics reassures lis-
teners of a very simple, down-home quality since they convey, after all, the
simple pangs of young love’ (Flinn 1992, 122). The song was plugged in
advance of the film’s release, the writers refusing to truncate it when the
publishers had baulked at its non-standard length, which would exceed the
normal limits of a piece of sheet music (Fordin 1996, 118). As in so many
Golden Age Hollywood products, the film’s considerable artistry was made
to appear effortless, an obvious example being the use of playback tech-
niques, which permitted singers to lip-synch without need for the facial
contortions necessary in real singing. Like Kurt London, Balázs lamented
the facial contortions of singers, never intended to be viewed in the merci-
less close-up detail beloved of the movie and television camera, and noted
that playback techniques had improved the situation (Balázs 1953, 277–8).
Such apparently easily won perfection helped promote the utopian feeling
at the heart of this quintessentially feel-good genre – a process particularly
noticeable in Garland’s two Christmas songs ‘You and I’ (Freed/Brown) and
‘Have Yourself a Merry Little Christmas’ (Martin/Blane).
A rash of period-set folk musicals followed in the wake of Meet Me
in St Louis, including Fox’s State Fair (dir. Walter Lang, 1945; songs by
Rodgers and Hammerstein; remade 1962) and Centennial Summer (dir.
Otto Preminger, 1946; songs by Kern and Hammerstein), MGM’s Summer
Holiday (see above) and Universal’s Meet Me at the Fair (dir. Douglas Sirk,
1952; music from various sources). The Oklahoma! tradition continued in
such musical westerns as MGM’s Annie Get Your Gun (dir. George Sidney,
1950; songs by Berlin), Seven Brides for Seven Brothers (dir. Stanley Donen,
1954; songs by Johnny Mercer and Gene de Paul) and Paramount’s Paint
Your Wagon (dir. Joshua Logan, 1969).
A crippling generic complacency was vividly demonstrated by Rodgers’
and Hammerstein’s impoverished attempts to transfer their stage hits to the
screen in a series of overblown and stiltedly theatrical widescreen produc-
tions for Fox: Oklahoma! (dir. Fred Zinnemann, 1955), Carousel (dir. Henry
King, 1956), The King and I (dir. Walter Lang, 1956) and South Pacific (dir.
Joshua Logan, 1958), with scores supervised variously by Alfred Newman,
choral arranger Ken Darby and Adolph Deutsch. Gerald Mast character-
izes these four films as ‘the personal revenge of Broadway creators on a
rival entertainment industry that had treated them shabbily two decades
earlier. The industry’s subtle retaliation was that these films barely lived
on the screen and do not survive appreciatively in critical repute’ (Block
2002, 101). Mast goes so far as to dismiss them as ‘opera films’ because they
constitute ‘reverential attempts, in a blockbuster era of Hollywood desper-

ation, to hang decorative sights on important music. They reject one of the
earliest discoveries of movie musicals – going back to 1929 – that space need
not remain constant while characters sing.’ He points out that in all but
Oklahoma! the camera does not for a single instant depart from the singer,
throwing emphasis on the song as song performance rather than plausi-
ble extension of character. The beach scenes in South Pacific remain sterile
stage routines in spite of a sumptuously realistic setting. Wooden actors
and characterless dubbed singing (the benefits of the then state-of-the-art
six-track stereo technology notwithstanding) sounded the death knell on
Fox’s stagebound ventures, at a time when no studio attempting a Broadway
transfer would dare risk alienating an audience familiar with a prior stage
production and its bestselling cast recording by indulging in even modest
creative experimentation.
The 1950s Fox extravaganzas, with their level of creativity in inverse
proportion to their budgets, were particularly disappointing since almost
two decades earlier MGM had shown just how magical and captivating a
truly imaginative musical screenplay could be. The Wizard of Oz (dir. Victor
Fleming, 1939), based on a fairy-tale story already treated on Broadway (in
1903) and in the medium of silent film (in 1910 and 1925), was the first
out-and-out fantasy film musical, receiving Academy Awards for both its
score and Harold Arlen’s song ‘Over the Rainbow’ – the latter famously
slated for jettisoning by studio moguls who considered it too dull. Musical
director Herbert Stothart, who provided sophisticated scoring for narrative
portions of the film including a memorably impressionistic soundscape for
Dorothy’s first tentative footsteps in Oz, commented two years later: ‘We
learned that a musical episode must be so presented as to motivate a detail
of the plot, and must become so vital to the story that it cannot be dispensed
with. The test today is – “If a song can be cut out of the musical, it doesn’t
belong in it”’ (quoted in Flinn 1992, 35). Although its profits were modest
on account of its huge budget, Oz secured MGM its position as the finest
and most prolific purveyor of original musicals in all subgenres until the
bottom dropped out of the market in the 1960s, and many of the studio’s
less ambitious projects in its heyday comfortably grossed four times their
budgets at the box office.
Key members of the Freed Unit at MGM responsible for this string of
successes were staff musicians who combined creative talent with adminis-
trative acumen, among them Roger Edens (both vocal arranger and associate
producer), Kay Thompson and Lela Simone; arrangers, composers, orches-
trators and musical directors including Georgie Stoll, Conrad Salinger and
the teenage prodigy André Previn; and front-rank directors like Mamoulian
(his last foray into the genre being the unremarkable Silk Stockings in 1957)
and the ubiquitous Minnelli. The on-screen talent numbered Judy Garland
(until her tantrums and absences from shooting grew to such unaccept-
able levels that her contracts were suspended in 1949–50), Mickey Rooney,
Cyd Charisse and Gene Kelly, the last revolutionizing the role of solo dance
in the genre (see below). Johnny Green assumed full executive powers as
studio music head in 1949 and claimed only Alfred Newman and Ray
Heindorf rivalled him in influence at their respective studios (Fox and
Warners). Green’s formidably controlling personality mistrusted the infor-
mal family atmosphere of the Freed Unit, which habitually rebelled against
his weekly departmental administrative meetings at which musicians of the
stature of Rózsa, Kaper, Raksin and Deutsch were compelled to sit at under-
sized school desks in order to listen to administrative minutiae (Fordin
1996, 301–2). Freed left MGM in 1970, by which time its fortunes had
seriously diminished: when executive James T. Aubrey was appointed to
overhaul the studio in 1969, he ordered the destruction of most of its music
library, including huge amounts of irreplaceable material in both notated
and recorded form. Four years later the studio ceased production and dis-
tribution of motion pictures, opting to make its money through television
series and gaining further income from commercial ventures such as its own
hotel in Las Vegas (Fordin 1996, 524).
Even more than in other Hollywood genres, and reflecting a trend that
persists to the present day, much of the orchestration and other aspects of
musical arrangements appeared needlessly overdone – even to some profes-
sionals in the thick of such work. The score for Fox’s Carousel transfer was
not atypical in involving the work of no fewer than six orchestrators, includ-
ing Earle Hagen, Nelson Riddle and Herbert Spencer. At MGM, Deutsch
attempted to make a stand on what he perceived as a pressing need for
simple orchestration: ‘At the first production meeting [in 1950] I said that
I would only do Show Boat if I could approach Kern’s music simply, as he
intended it to be heard and played’. (Kern was notorious for his abhorrence
of jazzed-up versions of his tunes.) Deutsch was obliquely referring to the
efforts of his colleague Salinger, who had worked on the Kern biopic Till
the Clouds Roll By (1947), about which Deutsch was ‘a little uncomfortable
because the arrangements, vocal as well as instrumental, were, I thought, a
little overembellished and overranged for a man who was as simple as Jerome
Kern’ (Fordin 1996, 336–7). Revealingly, when Previn was first given sole
responsibility for the scoring of an MGM musical in 1955 (It’s Always Fair
Weather, dir. Kelly and Donen) – which he recalled was the first time a single
musician and not a team of composers and arrangers had been responsible
for all aspects of the score from inception to recording – he still found the
arranging instincts of his apprenticeship got the better of him, admitting ‘I
don’t think that too many of the songs were very good and that’s because I
was too intent on having them sound clever or well arranged’ (Fordin 1996,
435). Nevertheless, the process of arrangement could produce understated
poetry when the art was at its finest: Salinger’s orchestration in the lyrical
moments of Meet Me in St Louis, for example, was economical and beauti-
fully executed in performance by Stoll, who brought forth an extraordinary
warmth of string tone and sensitive rubato from his players.
Elaborate dance routines had always been just as important in some
musicals as song, not only dances performed by massed forces – whether
stomping folk dancers in the old west or Berkeley’s psychedelically swirling
chorines – but those conceived as an expression of a character’s individu-
ality. In the mid-1930s Berkeley demonstrated how dance might spectacu-
larly release the visual images from their primary need to serve the diegesis
(Altman 1987, 70), but did not exploit the medium as an agent of character
or narrative. Contemporaneously, Fred Astaire and Ginger Rogers showed
in a series of virtuosic films for RKO how an integrated use of dance could
articulate romantic plots in a dynamic and thoroughly modern manner:
solo dance becomes self-expression, and the dancing couple dramatizes the
shift towards inevitable romantic union. Astaire and Rogers first appeared
together in Flying Down to Rio (dir. Thornton Freeland, 1933) and then
made The Gay Divorcee (dir. Mark Sandrich, 1934), which received the first
Academy Award for Best Song (Con Conrad’s and Herb Magidson’s ‘The
Continental’, forming the basis for the film’s danced climax). The pair went
on to showcase the songs of Berlin in Top Hat (dir. Sandrich, 1935), of Kern
in Swing Time (dir. George Stevens, 1936) and of the Gershwins in Shall We
Dance (dir. Sandrich, 1937). The scores for Astaire’s and Rogers’ first three
films were supervised by Max Steiner before his move to Warner Bros. and
replacement by Nathaniel Shilkret.
On the one hand, Astaire’s elegant and sophisticated dance routines
continued to promote the aura of utopian romance central to the genre
as a whole. As Altman puts it, ‘Time and again we watch Astaire and his
various partners fall in love only when they step into that privileged land
of make-believe created by dance. If lovers’ hearts beat in time to each
other, then the dance provides an opportunity to rehearse that rhythm, to
slide imperceptibly from indifference to passion’ (Altman 1987, 85). This
was only to be expected in the fairy-tale subgenre, which since Lehár’s The
Merry Widow of 1905 (filmed by Lubitsch for MGM in 1934), with its
apotheosis of the waltz, had equated dance with love. In Astaire’s partner-
ship with Rogers, however, the romantic interest was sharpened with an
innovative love–hate flavour, their dances sometimes beginning as bicker-
ing then moving towards union (Altman 1987, 161). Furthermore, Astaire’s
skills in tap-dancing were directly and symbolically pitted against Rogers’
classical dancing, nowhere more so than in Shall We Dance where the entire
plot revolves around his attempt to entice her into a vernacular mode of
physical expression (Braudy and Cohen 1999, 621–5). The film’s closing bal-
let parodies the impersonality of Berkeley’s massed ranks of faceless chorines
by making Astaire search out the true Rogers from a bevy of dancers masked
with two-dimensional reproductions of her face, the parody all the more
telling since Rogers had herself previously appeared in 42nd Street and Gold
Diggers of 1933. Berkeley’s choreography was also parodied alongside spoof
examples of the operetta style in the Marx Brothers’ Duck Soup (dir. Leo
McCarey, 1933), which muddles up idioms as diverse as spirituals, ragtime
and hillbilly music in its performative ‘orgy of chaos’ (Marshall and Stilwell
2000, 48) when war is declared in the Ruritanian state of Freedonia – the
name punning on Fredonia, NY, whence the brothers fled in disgrace after
a disastrous performance at the local opera house in their early vaudeville
career.
Gershwin’s music for Shall We Dance came at the end of his ambiva-
lent relationship with Hollywood, which began with Delicious (dir. David
Butler, 1931) and intensified when he moved to the West Coast following
the East Coast launch of Porgy and Bess in 1935, which had been directed by
Mamoulian and featured Eva Jessye’s choir. Porgy, renowned as an inven-
tive and satisfying treatment of African American subject-matter (albeit
filtered through the sophisticated uptown musical sensibilities identified by
Gabbard in Cabin in the Sky) and filmed by Preminger for Columbia in 1959,
had more than any other of his works reflected Gershwin’s serious artistic
aspirations; consequently, his reputation as a popular tunesmith seemed in
danger of dipping, and when about to commit himself to Hollywood in
1936 he cabled his agent to say in his own defence: ‘Am not highbrow. Have
written hits before and expect to write them again’ (Wyatt and Johnson
2004, 240). In the event, Shall We Dance elicited some of his most endur-
ing standards (‘They Can’t Take That Away from Me’, ‘They All Laughed’
and ‘Let’s Call the Whole Thing Off’) and the composer benefited from the
usual song-plugging in advance of the film’s release, telling a correspondent
in April 1937 that the finished product ‘is practically ready for public gaze
and if you turn on your radio you will hear the songs from it achieving a
rather quick popularity’ (Wyatt and Johnson 2004, 259). In spite of failing
health, part of his therapy for which was playing tennis with Schoenberg
in Beverly Hills, Gershwin also contributed music to RKO’s A Damsel in
Distress (dir. Stevens, 1937), starring Astaire but this time alongside Joan
Fontaine, and including the songs ‘A Foggy Day’ and ‘Nice Work if You Can
Get It’. In a common move, MGM later bought the rights to several Gersh-
win titles (e.g. Strike Up the Band in 1940 and Lady Be Good in 1941), using
them as crowd-pullers for their films though only utilizing a smattering of
Gershwin’s music in the course of the relevant picture. In 1943 the studio
made a complete version of Gershwin’s 1930 Broadway hit Girl Crazy (dir.
Berkeley and Norman Taurog), throwing in ‘Fascinating Rhythm’ (from
Lady Be Good) for good measure.
At MGM, Gene Kelly’s blend of virtuosic exhibitionism in dance rou-
tines and childlike immaturity of screen persona was very different in effect
from the polished sophistication of Astaire. Leo Braudy has summarized the
contrast between the two in terms which again suggest the tension between
theatricality and realism at the heart of so much cinema:
Astaire is the consummate theatrical dancer, while Kelly is more interested

in the life outside the proscenium. The energy that Astaire defines within a
theatrical and socially formal framework Kelly takes outside, into a world
somewhat more ‘real’ (that is, similar to the world of the audience) and
therefore more recalcitrant . . .
Astaire may mock social forms for their rigidity, but Kelly tries to explode
them. Astaire purifies the relation between individual energy and stylized
form, whereas Kelly tries to find a new form that will give his energy more
play. Astaire dances onstage or in a room, expanding but still maintaining
the idea of enclosure and theater; Kelly dances on streets, on the roofs of
cars, on tables, in general bringing the power of dance to bear on a world
that would ordinarily seem to exclude it. (Braudy and Cohen 1999, 624)
Kelly’s ascendency in song and dance coincided with the development of the
interpolated ballet sequence in film musicals, a natural consequence of the
dream sequences and Berkeley fantasy inserts in the 1930s. At the hands of
Minnelli, these increasingly elaborate ballets often took the form of a basic
psychoanalytical probing of the protagonists’ diegetic preoccupations (Mar-
shall and Stilwell 2000, 31). Among the choreographers influential on this
trend was Agnes de Mille who, amongst her many other achievements, had
choreographed the Oklahoma! première and was responsible for Louise’s
dream ballet in Carousel. Her comments on the need for a modern and
vernacular dance style instead of outworn (and therefore inexpressive) clas-
sical movements in essence parallels the relationship between the up-to-date
jazzy style of music that conflicted with doggedly persisting operetta modes
of composition:
Ballet gesture up to now had always been based on the classic technique and
whatever deviated from this occurred only in comedy caricatures. The style
throughout, the body stance, the walk, the run, the dynamic attack, the
tension and controls, were balletic even when national folk dances were
incorporated into the choreography.
We were trying to diversify the root impulse and just as Gershwin
impressed on the main line of musical development characteristics natural
to his own unclassical environment, we were adding gestures and rhythms
we had grown up with, using them seriously and without condescension for
the first time. This is not a triviality; it is the seed and base of the whole
choreographic organization. If dance gesture means anything, it means the
life behind the movement. (De Mille 1952, 307)
De Mille believed that each gesture should be a natural extension of the

performer’s assumed character, and that therefore the distinction between
acting and dancing was being eroded.
At MGM substantial balletic set-pieces, from which dialogue and song
absented themselves in order to foreground symbolic dance and elaborate
orchestral music, became a house speciality; but their increasingly elaborate
nature and length led to accusations of pretentiousness from critics who
felt the genre was here over-reaching itself. Examples include the surreal
ballet for Astaire and Lucille Bremer in Yolanda and the Thief (dir. Minnelli,
1945), choreographed by Eugene Loring; Kelly’s choreography for The Pirate
(dir. Minnelli, 1947), starring Garland and Kelly and featuring songs by
Cole Porter; the ‘Slaughter on Tenth Avenue’ ballet in the Rodgers–Hart
biopic Words and Music (dir. Norman Taurog, 1948); and the disconcerting
juxtaposition of vivid location shooting and the stylized studio ‘Turnstiles’
and ‘A Day in New York’ ballets in the reworking of Leonard Bernstein’s On
the Town (dir. Kelly and Donen, 1949). The climaxes of Kelly’s MGM dancing
career came with An American in Paris (dir. Minnelli, 1951), Invitation to
the Dance (dir. Kelly, 1952) and Singin’ in the Rain (dir. Kelly and Donen,
1952). Reverentially regarded by cineastes as diverse as Truffaut and Resnais,
and even by trenchant critic Pauline Kael, Singin’ in the Rain celebrates the
heady years of Hollywood’s transition to sound, so it was appropriate to
have Freed’s 1929 hit song at its heart, now transformed by the memorable
addition of Edens’s ‘chic, riff-like fill . . . constantly dancing rings around
its harmonic progression’ (Banfield 1998, 317). The plot also permitted the
film to celebrate the backstage tradition from a sophisticated perspective
that drew on the studio’s long experience of various other generic subtypes,
and a focal point is a fifteen-minute ‘Broadway Ballet’.
The seventeen-minute ballet sequence in An American in Paris, which
received its own main-title card, was a flamboyant interpretation of a man-
gled version of Gershwin’s symphonic poem in which a black-and-white-
clad Kelly cavorts amidst the saturated colours of a theatricalized French
capital, and is middlebrow fayre par excellence. More telling was Kelly’s
artistry in the film’s intimate moments, whether joining local kids in a
stichomythical interpretation of ‘I Got Rhythm’ or his wonderfully under-
stated nocturnal courtship dance with Leslie Caron by the banks of the
Seine, which rivals Astaire for its indissoluble link between technique and
expression. Even the virtuosic tap routine performed by Kelly atop the piano
in Oscar Levant’s garret – the extrovert modern dancer throughout the film
pointedly contrasted to the depressive and egotistical unemployed classical

musician portrayed by Levant – seems to grow naturally from the diegesis.
The film, which (like On the Town) hedged its aesthetic bets by juxtaposing
naturalistic location shooting and highly stylized theatrical sets, won no
fewer than six Academy Awards, including one for the arranging skills of
Saul Chaplin and Johnny Green.
The ‘Circus Ballet’ in Invitation to the Dance ran for half an hour of
screen time. Made at Elstree studios in the UK, the sequence had a score
commissioned from Jacques Ibert and recorded by the Royal Philharmonic
Orchestra in primitive circumstances and with inadequate equipment that
shocked both musicians and producers, the project progressing even less
auspiciously when Ibert had to interrupt work owing to the death of his
daughter in mysterious circumstances (Fordin 1996, 379–6). To add to
Kelly’s headaches, English composer Malcolm Arnold’s score for the ‘Ring
Around the Rosy’ sequence was rejected, even though it had already been
recorded, and was replaced by Previn (here undertaking his first compos-
ing assignment for the studio and being presented with the daunting task of
having to write ballet music to a predetermined image track). Another com-
poser working in the UK, Canadian-born Robert Farnon (later well known
for his grimly defiant signature tunes to the BBC’s 1970s wartime dramas
Colditz and Secret Army) was asked to provide music for a further sequence
destined for the cutting-room floor. The wary Lela Simone wrote to Freed
from England: ‘I am sticking with [Farnon] very closely so that we get an
American song sequence and not polite, English jazz. I have played a lot of
tracks from our pictures for him; this will give him an idea as to what we
want to hear’ (Fordin 1996, 387). A highlight of the film, demonstrating the
close links between dance and the audio-visual rhythms of the cartoon, was
the 22-minute ‘Sinbad the Sailor’ ballet with music by Rimsky-Korsakov
orchestrated by Salinger and conducted by Green, and animation by the
capable team of Hanna, Barbera and Quimby. This over-ambitious film was
put on ice and not released until 1957, by which time its arty nature coin-
cided with the general slump in the industry and it consequently made a
loss. Greater contemporary resonance was achieved by the jazz-dance style
of United Artists’ West Side Story (dir. Robert Wise and Jerome Robbins,
1961), of which the stereo soundtrack album long retained a record-breaking
place at the top of the Billboard album charts (Banfield 1993, 39), though the
distinctively energetic choreography had already been a prominent element
in the stage show and the theatrical conception of the work now clashed
somewhat against the cinematographic realism of the film version.
As Heather Laing has argued, formal song types in the musical transcend
their structural limitations and enhance the narrative even as they appear
to interrupt it:
While everything else is undergoing radical disruption, the music is at its

most contained, formulaic and, by extension, familiar and safe. The use of
song is bound to make the audience more aware of its presence as music,
potentially reducing some of the unique extremes of its narrative power.
Apart from the pleasure in the familiarity of the form, however, the
presentation of the number actually counteracts this possibility of narrative
weakening, by concentrating the credit for excessive intensity, immediacy,
and emotional depth, away from the actual music, and onto the performing
character. (Marshall and Stilwell 2000, 10–11)
This musical emphasis on character is naturally most effective when the per-
former is blessed with both a powerful personality and a singing voice that
is not (or at least does not obviously appear to be) dubbed. Not surprisingly,
therefore, as the musical emerged from its decline in fortunes in the later
1950s its sporadic continuation depended heavily on the star quality of its
performers. Some reinvented themselves. Astaire’s later career, for example,
replaced his pre-war image of modernity and youth into its precise opposite:
in the 1950s he was often portrayed as the older mentor of younger women
to whom he presented an image of conservative maturity (Cohan 2002, 95).
Allied to phenomenal record sales and enormous sex appeal, Elvis Presley’s
charisma certainly accounts for the success of his films, dismissed by Altman
as ‘poorly made in every way’ (Altman 1987, 194); other popular perform-
ers had discovered in the show subgenre a convenient way to disseminate
their modern styles on screen, a process most spectacularly shown by the
international success of Bill Haley and the Comets in Rock Around the Clock
(dir. Fred F. Sears, 1956), and emulated by emerging pop stars in the UK
(see Chapter 10).
A more chaste newcomer in the traditional fairy-tale genre was Julie
Andrews, who gave brilliantly fresh and direct performances in Disney’s
Mary Poppins (dir. Robert Stevenson, 1964) and Fox’s The Sound of Music
(dir. Robert Wise, 1965), showing what Peter Kemp has identified as a
refreshing commitment to the narrative importance of songs rather than
merely treating them as vehicles for display (Marshall and Stilwell 2000, 60).
The Sound of Music, dating from after Hammerstein’s death, was far more
compelling than the earlier Rodgers–Hammerstein screen transfers; unlike
the lifeless ocean in South Pacific, the stunning Alpine scenery here almost
becomes a character in the drama. Contemporaneous hit stage transfers
featuring British talents were My Fair Lady (dir. George Cukor, 1964) and
the multiple Oscar-winning Oliver! (dir. Carol Reed, 1968). Flamboyant
charisma and powerful musical and creative talents were the hallmarks of
Barbra Streisand, whose personality stamped its indelible presence on Funny
Girl (dir. William Wyler, 1968) and its sequel Funny Lady (dir. Herbert Ross,
1975), and her idiosyncratically self-authored drama Yentl (dir. Streisand,
1983), the last with Oscar-winning music by Michel Legrand (whose Franco-
phone film musicals are discussed in Chapter 8). When Streisand performed
numbers live on set in Funny Girl, her accompaniment was fed to her via
a small speaker which had to be kept from the camera’s view, necessitating
close-up photography (LoBrutto 1994, 7).
The fantasy musical had passed squarely to Disney, since animation
not only aided suspension of disbelief in an increasingly cynical world but
proved to be an ideal medium in which to perpetuate the genre’s longstand-
ing association with wholesome and uncontroversial family entertainment.
Among Disney’s roster of fairy-tale musical features after the groundbreak-
ing Snow White and the Seven Dwarfs in 1937 (to which The Wizard of Oz had
consciously set itself up as a live-action rival two years later) were Cinderella
(1950), The Lady and the Tramp (1955), The Sword in the Stone (1963) and
The Jungle Book (1967). The studio’s commitment to the musical continued
until the end of the century and beyond, with a canny use of up-to-date
pop idioms aiding the success of more recent ventures such as The Little
Mermaid (1989), Beauty and the Beast (1991), Aladdin (1992), The Hunch-
back of Notre Dame (1996) and Hercules (1997), all featuring the music of
Alan Menken, and The Lion King (1994), including the voice of Elton John.
Banfield notes that in several places the music for Beauty and the Beast pays
direct homage to the operetta tradition, and this may in part account for
the successful reverse-transfer process which put the film musical on the
Broadway stage in 1994 (Banfield 1998, 333). A stage version of The Lion
King, premièred in Minneapolis in 1997 and directed by Julie Taymor, also
met with critical acclaim and popular success. That these transfers from
screen to stage could equally prove to be commercially disastrous, however,
was shown by the Broadway staging of Gigi which bombed in 1973 (Fordin
1996, 495), even though MGM’s original 1958 film had broken all previous
records for the genre by receiving nine Academy Awards, including Best
Picture, Best Song, Best Director (Minnelli) and Best Scoring (Previn). A
1989 Broadway production of Meet Me in St Louis similarly flopped (Steyn
1997, 264).
In contrast to unthreatening familial conservatism, the later film musical
found a particular niche as an expression of camp. Amongst the stars who
were reincarnated after their early success, Garland was transformed from
the girl next door into a camp icon after her androgynous ‘Couple of Swells’
routine in Irving Berlin’s Easter Parade (dir. Charles Walters, 1948). Later
film musicals turned camp to very different ends, whether as an expression
of a regional gay culture such as that in Spain (see José Arroyo in Marshall
and Stilwell 2000, 70–9) or as a darker dramatic agent in adult reworkings
of the backstage subgenre. Most flamboyant of all was Tim Curry’s extraor-
dinary performance in the travesty role around which the screen transfer of
Richard O’Brien’s The Rocky Horror Picture Show (dir. Jim Sharman, 1975)
revolves. This musical was significant not only for its cross-generic bor-
rowings from science-fiction and horror, but also for the way in which its
filmic version only achieved belated cult status through late-night showings
involving a direct audience participation that added a further voyeuristic
level to that of the on-screen onlookers who formed an essential part of the
diegesis. Subtler and more disturbing was the role of the MC in Cabaret
(dir. Bob Fosse, 1972; songs by John Kander and Fred Ebb). Another stage
transfer, but effected with a keen eye for idiomatic cinematography and
sometimes powerfully symbolic montage (with parallel editing, for exam-
ple, used to link a diegetic Tyrolean slapping dance with the brutality of
Nazi brownshirts in order to suggest a link between hedonism and sadism),
the film recounts its central bisexual love triangle in a period backstage
setting inevitably tapping memories of that finest of all early German film
musicals, The Blue Angel (see Chapter 2), aided by the dynastic star quality
of Liza Minnelli, charismatic daughter of Garland and Minnelli. As Mark
Steyn notes, Fosse ‘shrewdly eliminated all the “book numbers” . . . and kept
only the “real” songs, performed in the Kit Kat Klub; Fosse understood that
young moviegoers no longer accepted the musical’s defining convention –
that a guy could walk down the street and burst into song with full orches-
tral accompaniment’ (Steyn 1997, 214–15). A few years later Liza Minnelli
joined Robert De Niro in Scorsese’s New York, New York (1977), an attempt
to update the musical through the medium of jazz: influenced by the work
of John Cassavetes (see Chapter 5), Scorsese included scenes improvised
by the actors in a structural debt to the aesthetic of jazz extemporization.
Following the example of Vincente Minnelli, he developed a method of edit-
ing camera shots in accordance with the lengths of musical phrases, such
carefully rhythmicized montage also retained in his later non-musical films
(Thompson and Christie 1989, 69).
Also in the 1970s, two trends emerged that gave the film musical a new
lease of life without significant impact on its general conventions. First,
flamboyant and energetic rock operas transferred from stage to screen with
some success, notably Andrew Lloyd Webber’s Jesus Christ Superstar (dir.
Norman Jewison, 1973) and The Who’s Tommy (dir. Ken Russell, 1975).
Second, a new pretext for featured song and dance was launched with the
dance musical, in which romantic, social and even period settings were all
geared squarely towards a youth audience, and dance re-emerged as a vehicle
for self-expression. This movement was spearheaded by the disco milieu of
Saturday Night Fever (dir. John Badham, 1977), the phenomenal soundtrack
sales of which are examined in Chapter 10; its star John Travolta went on to
appear in the commercially successful screen transfer of the more amiable
Grease (dir. Randal Kleiser, 1978), and later examples of the dance-oriented
genre included Fame (dir. Alan Parker, 1980), Flashdance (dir. Adrian Lyne,
1983) and Dirty Dancing (dir. Emile Ardolino, 1987), the formula persisting
into the new century with Save the Last Dance (dir. Thomas Carter, 2001).
Disco and dance scenes featuring dialogue were generally shot with the aid
of a subsonic ‘thumper’ beat, sounding below the lowest pitch range of dia-
logue and not therefore interfering with its recording: ‘You play the music
with the thumper’, commented production sound mixer Les Lazarowitz,
‘drop out the music, and you just leave the thumper going – it gets the
crowd going’; to minimize noise further, dancers sometimes performed in
bare feet or on carpets (both kept carefully out of the camera’s field of
vision), and principal actors received the music on earphones to aid syn-
chronization while the remaining dancers performed without it (LoBrutto
1994, 122–3).
In the late 1980s Altman was pessimistic about the film musical’s future:
‘Down one path lies the death of the musical by subservience first to Broad-
way and then to the recording industry; down the other lies the death of the
musical by self-inflicted wounds. Today we retain only a limited produc-
tion of children’s musicals (usually cartoons), adolescent musicals (usually
dance fad or concert-oriented), and old folk musicals (nostalgia compila-
tions or throwbacks)’; he lamented that visionary maverick directors like
Fosse remained few and far between (Altman 1987, 121). The situation had
not changed two decades later, with the sporadic appearance of stage trans-
fers such as Evita (dir. Alan Parker, 1996), Chicago (dir. Rob Marshall, 2002)
and Rent (dir. Chris Columbus, 2005), a steady trickle of (now mostly CGI)
animated musicals in the Disney tradition, and the very occasional burst of
originality from a bold director. Two entirely different examples of the latter
are the by turns realistic and surreal irruptions of song in the narrative of
the Homeric jailbreak comedy Oh Brother, Where Art Thou? (dir. Joel and
Ethan Coen, 2000) and the kaleidoscopic audio-visual exuberance of Baz
Luhrmann’s Moulin Rouge! (2001) – the latter a heady MTV-style updating
of various generic formulae with wildly anachronistic music capitalizing on
contemporary pop hits. This triumph of fashionable style over banal content
retained a strong debt to its forebears, including backstage elements in its
musical-within-a-musical, the gesture now acknowledging the global village
in its inclusion of a Bollywood routine. The main strength of Moulin Rouge!
lies in its vivid sense of fantasy and sometimes lurid stylization, its original
screenplay light-years away from problematic modern stage transfers with
their tendency to abandon theatricality in favour of detailed realism (Evita)
or package dramatic stylization into specific numbers rather than the whole
(Chicago). A further and significant glimmer of hope for the genre came in
2007 with Tim Burton’s stunningly imaginative (and hideously gory) inter-
pretation of Stephen Sondheim’s Sweeney Todd: The Demon Barber of Fleet
Street, a film distinguished not only by its perfect wedding of nightmarish

Gothic fantasy with stylized theatricality but also by its bold initiative in
using only the actors’ natural singing voices on its soundtrack; Sondheim’s
characteristic skill in moving seamlessly from speech to arioso and thence
to fully blown song was a major factor in the venture’s success.
Perhaps the most important aspect of the musical in terms of the general
history of film music is the genre’s bold breaking down of the distinction
between the diegetic and nondiegetic, which ‘blurs the borders between the
real and the ideal’ (Altman 1987, 63). Crucially, the idealized music tends to
predominate, being pre-recorded and frequently dictating the course of the
narrative, with ambient and other diegetic sounds suppressed to create what
Altman terms a ‘supra-diegetic’ music: in this context strictly diegetic music
functions as a bridge between a music-less reality and the superabundance
of music in the idealized realm, permitting an ‘audio dissolve’ between the
constituent sonic planes. Rhythmicized bodily motion can also effect this
transition, for example Astaire’s metrical walking allowing him to move
seamlessly from conversation to song and dance, the rhythmic organization
of quotidian physical activities in Oklahoma! such that ‘The rhythm of life
already constitutes a dance’ (Altman 1987, 307), the rhythmicized dialogue
matching train noises and the paralleling of gossips with clucking hens in
The Music Man (dir. Morton Da Costa, 1962), and the numerous examples
where characters hum tunes to themselves as a seemingly natural mechanism
for the implausible calling forth of an invisible orchestra (Altman 1987,
67–8).
Scoring Shakespeare
In an age increasingly suspicious of the concept of literary and musical
canons, and the cultural elitism inevitably engendered by them, film ver-
sions of William Shakespeare’s plays no longer need to be defended from
accusations of populism. At their time of writing, in any case, Shakespeare’s
plays were a prime example of commercial art, and depended entirely upon
popular success for their survival in the playhouse repertoire; and there has
been no more commercial an art than cinema, which has been responsible
for disseminating the bard’s work to a much wider audience than would have
been possible through the medium of live theatrical performance. Never-
theless, some might choose to agree with Erwin Panofsky that Shakespeare
in the cinema will still receive its highest accolades from those ‘not quite
in sympathy with either the movies au naturel or Shakespeare au naturel’
(Braudy and Cohen 1999, 284), even if the international power of cinema
is such as to have entirely altered our collective awareness of how certain
of Shakespeare’s filmed plays might otherwise be interpreted (Hammond

1981, 71).
In addition to illustrating various approaches to the task of making
modern adaptations of ‘classic’ texts, and providing representative and con-
trasting examples of the achievements of some of the foremost composers
who have worked for the cinema, filmed Shakespeare (as do filmed opera
and filmed musicals) generates discussion of the general problems inherent
in creating screenplays from theatrical works: for example, the essential dif-
ference between the framing device of the theatre’s proscenium arch and the
implied continuation of diegetic space beyond the borders of the cinema
screen; the fact that the camera watches the action on behalf of the spectator,
picking and choosing points of view that in the live theatre the spectator can
amend according to personal curiosity; the challenge of finding a workable
balance between realism and theatrical stylization; and the need to make
both individual visual images and narrative method cinematically viable,
often involving techniques such as the use of flashback to illustrate a nar-
rated event (e.g. the drowning of Ophelia in Hamlet). This last device has
been widespread, resulting from a ‘modern tendency to add substance to
what Shakespeare leaves shadowy’ (Craik 1995, 91), though it may be noted
that, as early as Colley Cibber’s production of Richard III in 1700, the young
princes were murdered on stage in order to increase the element of violence
in the play, this being contrary to Shakespeare’s evidently deliberate strategy
of keeping all murders unseen until the death of the title character himself
(Hammond 1981, 69, 98).
As far as production styles are concerned, film directors are given gen-
erous leeway on account of the paucity of specific stage directions in Shake-
speare’s texts. This freedom has been fully explored in modern Shakespeare
stagings in the theatre, and the fact that productions of the plays were
anachronistic even in Elizabethan times has been a justification for attempts
to modernize settings in some film treatments. Music for Shakespearean
films has tended to follow conventions well established in other genres, with
mixed results, and Michael Hattaway has noted that ‘poetical or lyric pas-
sages are deemed to need music to signal their “abnormality”’ (R. Jackson
2000, 90), giving as an example the treatment of ‘She Never Told Her Love’
from Trevor Nunn’s Twelfth Night (1995; music by Shaun Davey), intercut
with images of crashing waves.
Shakespearean plots were a popular subject for magic-lantern presenta-
tions in the nineteenth century, and silent film continued the tradition by
constructing animated tableaux based on the bard’s plays. One of the first
silent Shakespeare films was a five-minute Hamlet shown at the 1900 Paris
Exposition and starring Sarah Bernhardt in the title role (R. Jackson 2000,
117). As we have seen, a tightening of copyright legislation in 1908 led to
a rash of film productions based on non-copyright literature and opera,

and Shakespeare’s plays thus became a favourite source for single-reel silent
films. These pocket-sized productions – a representative selection of which
was released on DVD by the British Film Institute in 2004, with new music
commissioned from Laura Rossi – were originally screened with the same
stereotyped music as other silent films, and featured the exaggerated acting
(designed to be effective when viewed from a distance in a live theatrical
production but inappropriate to the intimacy of the screen) typical of the
film d’art movement. Among the more enterprising early attempts to unite
the silent image with an element of sound was actor Frederick Warde’s tex-
tual recitations accompanying a 1913 silent film of Richard III, the earliest
extant full-length American feature film (R. Jackson 2000, 47). In the UK, a
silent version of A Midsummer Night’s Dream featured music compiled by
prolific film arranger Hans May (Huntley [1947], 28).
With the advent of the sound film and its capability for recording
Shakespeare’s words, fundamental critical issues concerning the propriety
of screen adaptations began to come into play. At the same time, produc-
tions of Shakespeare had moved increasingly away from nineteenth-century
romantic values. For example, as early as 1901 a seminally minimalist pro-
duction of Henry V at the Shakespeare Memorial Theatre, Stratford-upon-
Avon, demonstrated the usefulness of continuous action, an innovation later
adopted as an essential feature of filmed Shakespeare (Craik 1995, 84). In
the 1930s, many commentators were disappointed by what they perceived
as the sound cinema’s reversion to the dated mannerisms and procedures of
the old actor-manager’s theatrical spectacle, involving not merely melodra-
matic styles of acting and pictorial set design, but a return to such perceived
inauthenticities as cavalier cutting of the text. In addition, cinematic real-
ism seemed at odds with the tendency towards austerity and stylization in
modern theatrical productions. A reliance on incidental music in a heavily
romantic style seemed distinctly old-fashioned – and, even worse to modern
theatrical sensibilities, background music might typically be used to bolster
dialogue.
These problems were aired after the appearance of the first treatment of
a Shakespeare play as a sound film: the opulent version of A Midsummer
Night’s Dream directed for Warner Bros. in 1935 by veteran Austrian stage
director Max Reinhardt, who had staged the play on Broadway. According
to one anonymous reviewer, it demonstrated ‘all the faults that grandiose
stage productions of Shakespeare once committed but have now happily
outgrown’ (The Times, 17 October 1935), while Mitry condemned its stage-
bound theatricality as an ‘absurd cardboard fairyland totally out of place
in the cinema’ (Mitry 1998, 323). The film, which utilized Mendelssohn’s
incidental music arranged by Korngold and featured fairies choreographed
in a manner recalling the great Berkeley musical routines, was dismissed

by Panofsky as ‘probably the most unfortunate major film ever produced’
(Braudy and Cohen 1999, 284). Kracauer took Allardyce Nicoll to task for
suggesting that the film aimed to rekindle the audience’s interest in Shake-
speare’s words through the use of sensuous visual stimuli, and criticized it
for failing to achieve the necessary balance between text and image (Weis
and Belton 1985, 129). Alberto Cavalcanti, lamenting in 1939 the fact that
film music in general still persisted in using outdated romantic styles, criti-
cized MGM’s rival treatment of Romeo and Juliet (dir. George Cukor, 1936;
mus. dir. Stothart) for both its highly romanticized production and reliance
on Tchaikovsky’s over-used melody. Cavalcanti noted that Tchaikovsky’s
music
fitted the production perfectly – that is to say, it was music of the indoors,
heavy with scent, unventilated, introverted, consorting well with the
glorified seraglio that was the set designer’s picture of ancient Verona . . .
This is the musical accompaniment, if you please, of a play by Shakespeare
which presents one of the purest love-stories of all time – full of stark, sharp
terrifying beauty . . . [I]n another production, I should certainly like to
entrust the music to a good modern composer. Shakespeare’s strangely
universal genius needs to be interpreted anew in every age – by the most
modern means. The recent film Romeo and Juliet was thirty years out of date
all the way through.
Not for the dignities of Shakespeare only, but also for all other dramatic
presentations, I plead for modern music, mood-music, because I am sure
that it has a great deal to contribute. (Weis and Belton 1985, 107)
One ‘good modern composer’ who soon made an international reputa-

tion for himself as a composer of fine Shakespearean film scores was William
Walton. His music for As You Like It (dir. Paul Czinner, 1936) was not entirely
successful in negotiating the usual problems attending the composition of
comedic scores, and has been criticized by Hattaway for ‘perfunctory and
incongruent set-pieces’ suggesting ‘that tone in a comedy of wit is probably
best modulated primarily through verbal delivery’ (R. Jackson 2000, 93).
The young Benjamin Britten wrote a review of Walton’s score for the journal
World Film News in April 1936 in which he described ‘the Grand Introduc-
tion over the credit titles – pompous and heraldic in the traditional manner’,
the use of Elizabethan songs, some effective dovetailing and restrained use
of a leitmotif – all of which would remain features of Walton’s later film
scores. At the same time, Britten felt that Walton had not always troubled to
adapt his scoring to the demands of recording technology: ‘One cannot feel
that the microphone has entered very deeply into Walton’s scoring soul. A
large orchestra in which strings are very prominent has been used, and in
the accompanying pastoral music one is conscious of the energetic ranks of
the London Philharmonic sweating away behind the three-ply trees’ (Kildea
2003, 21).
As You Like It starred Laurence Olivier as Orlando, and it was under
Olivier’s direction that Walton went on to achieve his most accomplished
Shakespearean scores. In 1944 they collaborated on Henry V, its vision of the
English victory at Agincourt carrying a clear wartime propaganda message:
in the main titles the film is dedicated ‘to the Commandos and Airborne
Troops of Great Britain, the spirit of whose ancestors it has been humbly
attempted to recapture’, and the project was partly conceived as a morale-
booster in the year of the Allied D-Day landings in Normandy. Its climactic
battle sequence drew heavily on both the visual imagery and musical style of
the Battle on the Ice from Eisenstein’s and Prokofiev’s equally propagandist
Alexander Nevsky (1938); the influence of D. W. Griffith’s epics and Errol
Flynn’s swashbucklers has also been noted (Craik 1995, 94). As befitted such
a prestigious production, the music budget was not abstemious and the total
cost of preparing the music track was in excess of £25,000 (Huntley [1947],
76). Olivier’s direction was fresh and original, with the film combining both
overtly theatrical presentation and exterior realism. Walton’s music enables
a smooth transition from the opening panoramic view of Elizabethan Lon-
don (accompanied by rousing music for chorus and orchestra in a modal
style redolent of Vaughan Williams) to the interior of the Globe Playhouse,
featuring pseudo-Elizabethan diegetic music, from where the scene is ‘trans-
ported’ to exterior locations in France via impressionistic underscoring as
the chorus/narrator describes the channel crossing by Henry’s army. The
substantial orchestral cues accompanying the Agincourt charge and battle,
which include reworkings of fanfares first heard diegetically in the Globe
Playhouse scene (where the diegetic music features tabor and harpsichord to
create a dash of authentic period colour), at times betray the clear influence
of Sibelius in their use of driving rhythmic patterns animating a funda-
mentally slow harmonic rhythm. The thrilling battle sequence (analysed in
detail, with photographic stills, in Manvell and Huntley 1957, 79–91) was
singled out by many as the high point of the score. Olivier himself pointed
out that Walton’s music was the only thing that made the charging French
knights seem plausible; in reality, they were Irish farmers riding their own
horses (Walton 1988, 94–5). It had originally been intended to pre-record
the music for this scene as a ‘guide track’ so that the live action could be
filmed in synchronization with it, and a piano reduction was recorded for
this purpose but, in the event, not used (Kennedy 1989, 123); as a result,
Walton declared the film ‘more of a bloody nuisance than it is possible to
believe . . . I seem to get no chance of settling down to the music & of course
there is going to be the usual hell of a rush’ (Hayes 2002, 147). With higher
spirits the composer had previously written to his assistant Roy Douglas to
ask ‘How does one distinguish between a crossbow & a long bow musically
speaking?’ (Hayes 2002, 145). In the event the archers were not furnished
with music, and praise was subsequently lavished on the editorial decision
not to underscore the sound of the first volley of English arrows, which
occurs immediately after a climax in the music. Other passages in Walton’s
score demonstrated an understated poignancy, notably the haunting pas-
sacaglia accompanying the death of Falstaff; the sombre ground bass is, in an
appealingly ironic gesture, derived from a traditional drinking song heard
in earlier comedic moments at the Boar’s Head inn. In the gently lyrical cues
for scenes featuring Katherine of Valois and the Duke of Burgundy, Walton
captures the fairy-tale quality of Olivier’s vision of the French landscape
with music reminiscent of Delius (and incorporating a French folksong).
As Anthony Davies has pointed out, the French are here not only seen as
an enemy: ‘the poised elegance of the frankly one-dimensional and stylised
castles and the landscape of France accompanied on the sound track by Wal-
ton’s wistful music . . . and finally Burgundy’s portrait of “this best garden of
the world” spoken again over Walton’s evocative music, all suggest qualities
of civilised life which the English need and for which they unconsciously
yearn’ (R. Jackson 2000, 169). Extracts from Walton’s score became popular
concert items after they were performed at the 1945 Promenade Concerts,
and the two movements for strings alone issued on a gramophone recording
under the composer’s baton in the same year.
The high point of the Olivier–Walton collaboration was Hamlet, released
in 1948 and shot in monochrome, which distinguished itself internationally
as the first non-American film to receive the Academy Award for Best Picture.
Although the costumes and settings were traditional enough, both produc-
tion design and cinematography were heavily steeped in the expressionistic
mannerisms of film noir (high-contrast lighting, low-level upward camera
shots, point-of-view camera roving restlessly down long corridors, beckon-
ing spiral staircases, and voice-overs in soliloquies); in this respect, and in
Walton’s avoidance of pomp-and-circumstance musical style in favour of
more atmospheric writing, the film offered a contemporary feeling to its
first viewers as a Shakespearean thriller (a whodunit, in fact, since Ham-
let sets out to expose the uncle who has murdered his father). Among the
score’s felicitous moments are the disconcerting harmonization of a diegetic
tolling bell with chords variously dissonant and concordant with it; an
intense chromatic fugato first representing ‘Something . . . rotten in the
state of Denmark’ and then, like the labyrinthine castle interiors, used to
portray Hamlet’s tortured psyche; irrational, non-functional harmony for
both the apparition of the ghost of Hamlet’s father and Ophelia’s descent
into madness; and the blurring of diegetic and nondiegetic music during the
play-within-the-play. In places, the orchestration is decidedly unorthodox,
as when the voice-over describing Ophelia’s suicide is backed by a recitative

for unaccompanied violins, and when the concluding funeral march (its
style appropriately reminiscent of Rózsa’s film noir scores) seems to fizzle
out on a high and greatly prolonged violin note before the final statement
of the tonic in the bass.
After this extraordinary achievement, the Olivier–Walton Richard III
(1955) came as a disappointment. In its way as patriotic in outlook as Henry
V, the film’s preoccupation with ‘The Crown of England’ (the final stark
caption of the main titles) was stimulated by undimmed national memories
of the Coronation of Queen Elizabeth II two years before; not surprisingly,
Walton drew heavily on the idiom of his own coronation marches to cre-
ate stirring music that now seems dated. Walton, who wrote the score in
haste, clearly responded with tongue in cheek to the director’s request for
ceremonial music, and added to the score of the Prelude the facetious mark-
ing ‘Con prosciuto, agnello e confitura di fragiole’ (‘with ham, lamb and
strawberry jam’: see Kennedy 1989, 194). Olivier’s film was a transfer of
an existing stage production, shot at speed with little understanding of the
technical difficulties of Paramount’s widescreen VistaVision; the film suc-
ceeded mostly on account of Olivier’s extraordinary acting, which included
addresses direct to camera. Walton’s music reuses some gestures from his
earlier scores (e.g. lyrical solo oboe and strings for both Hamlet’s Ophelia
and Richard’s Anne), is in places cloyingly sentimental (e.g. Clarence’s prayer
in the Tower), and frequently resorts to mickey-mousing more appropriate
in a Carry On comedy (e.g. Anne’s collapse on the stairs beneath Richard’s
throne, Richard’s prodding Buckingham in the chest and – most graphic of
all – Richard’s dying spasms and final collapse). Given the effectiveness of
the extraordinarily dissonant outburst as the deformed Richard stares with
cold menace at young York after the latter’s taunt (‘you should bear me on
your shoulders!’), it is regrettable that the score as a whole seems consis-
tently reluctant to engage with the powerful psychological undercurrents of
Shakespeare’s play.
Walton and Olivier were to have collaborated on a fourth Shakespeare
film – a version of Macbeth – but this project folded in 1958 owing to a
lack of financial backing. By this time Orson Welles had also applied his
directorial talents to Shakespeare, filming two plays, both with music by
European composers. His Macbeth (1948), like Olivier’s Hamlet (to which it
was unfavourably compared at the time), revealed both expressionistic and
theatrical influences, and was furnished with an eerie score by Jacques Ibert
(standing in for Herrmann, who had left the project) that included choral
breathing effects and impressionistic string harmonics for the supernatural
elements. When the film was restored to its original form in 1980, having
previously been dubbed with American accents and cut by Republic Pictures,
Ibert’s substantial overture was included (D. Cook 2004, 349). Between 1948
and 1952, Welles worked on Othello in Europe and this film, with music by
Francesco Lavagnino and Alberto Barberis, won the Grand Prix at the 1952
Cannes Film Festival; it was restored, with the orchestration reconstructed by
Michael Pendowski, in 1981 and re-released in 1992 to celebrate its fortieth
birthday, on which occasion the music track was digitally re-recorded. Welles
returned to Shakespearean subject-matter with Chimes at Midnight (1966),
a Spanish–Swiss co-production based on Henry IV and once again scored
by Lavagnino, whose tense modernism had contributed immeasurably to
the impact of Othello.
At the time when Miklós Rózsa’s attentions were increasingly engaged by
epic topics (see Chapter 5), he provided a score to Shakespeare’s Julius Cae-
sar (dir. Joseph Mankiewicz, 1953). As with his other epic scores, Rózsa
employed quartal harmony and organum passages to create an archaic
effect, especially in ceremonial fanfares, and – like Walton – had recourse
to the church modes, especially the Mixolydian, with its major tonality and
flattened seventh always making it a favourite for pseudo-archaic ceremo-
nial music. Perceptively, Rózsa refrained from providing music for Caesar’s
assassination since the event is portrayed as banal, and the perpetrators are
unable to grasp the enormity of their actions. From this point on, there is no
music for approximately half an hour: underscoring returns only when Mark
Antony incites the crowd to revenge. Drawing on Elizabethan raw materials,
Rózsa reworks John Dowland’s lute song ‘Now oh now I needs must part’,
which is first heard diegetically when performed by a young boy accompa-
nying himself on a lyre; it is immediately taken up in the underscore and is
much later developed in a sombre orchestral fantasia (as often in this com-
poser’s work, clearly influenced by the style of his compatriot Bartók) when
the shattered lyre is discovered by a centurion amongst post-battle debris.
Among the simple but effective stock devices used elsewhere in the score are
non-functional harmonies scored for string harmonics for the appearance
of Caesar’s ghost, and ritualistic drum ostinati lending excitement to the
climactic Battle of Philippi.
Zeffirelli made his first foray into filmed Shakespeare with The Taming of
the Shrew in 1966, casting Richard Burton and Elizabeth Taylor in the leading
roles for obvious commercial reasons. The score was composed by Nino
Rota, who balanced lyrical scoring with an archaic mood well suited to the
period décor; Hattaway has noted how the romantic theme at the opening
was given ‘an Elizabethan flavour’ when the scene moved to within the
walls of Padua (R. Jackson 2000, 93), a scheme recalling the similar balance
between Walton’s diegetic ‘Elizabethan’ music in the Globe Theatre scenes
of Henry V and more flamboyant orchestral accompaniment for exterior
scenes. It was the follow-up film of Romeo and Juliet in 1968, also with
music by Rota, that did most to establish Zeffirelli’s international reputation,

being spectacularly more successful than Renato Castellani’s 1954 film of
the same play (with music by Roman Vlad). Having taken the risk of casting
two unknown youngsters in the title roles, and against all predictions to
the contrary, Zeffirelli’s Romeo and Juliet netted two Academy Awards (for
costumes and cinematography) and grossed $50 million, injecting much-
needed funding into the then-ailing Paramount.
Rota’s score for Romeo and Juliet centred on a simple romantic love
theme, utterly Italianate in character with its hints of Verdi and Puccini and
clearly a strong influence on the melodic style of other Italian film com-
posers such as Ennio Morricone (see Chapter 9). Such was the popularity
of this theme that Capitol Records were persuaded to issue a soundtrack
album when none had originally been planned (M. Walker 1998, 172). The
theme first appears as a diegetic song, accompanied by flute and harp, then
grows increasingly intense in the underscore as an accompaniment to the
protagonists’ first kiss, when later they part company, as a prominent com-
panion to the preparations for Juliet’s feigned death, and at the climax of
the denouement in the mausoleum. Far from being a simple ‘theme’ score,
however, Rota’s music is elsewhere finely judged for dramatic effect. As in
The Taming of the Shrew, pseudo-Elizabethan style is employed under the
Prologue (spoken by Olivier as a voice-over to the opening shots of Verona),
and simple diegetic music with a period flavour is predominant – and care-
fully matched to the prevailing dramatic mood – until the love theme begins
to assert itself. Rota’s fine sense of pacing is evident in the first balcony scene:
simple lyricism underscores the intercut soliloquies of the lovers, ceasing
abruptly when they first address each other to allow the sharpness of the
dialogue full projection; music then returns as they exchange passionate
vows, thus framing the scene. A simple restatement of the love theme leads
into shots of Romeo rushing deliriously away, accompanied by frisky dance
music first heard diegetically earlier in the film. Following this sequence,
music is used sparingly, with the lovers’ kiss in Friar Laurence’s cell enacted
in silence, and virtually no music provided for the prolonged sword-fight
sequences apart from a simple contrapuntal lament for the death of Mer-
cutio and a brief burst of activity as Romeo pursues Tybalt.
By the time Zeffirelli tackled Hamlet in 1990, Rota (who had continued
to collaborate with the director on theatrical Shakespeare projects, including
in the 1970s an aborted plan to write a musical on Much Ado About Nothing)
was long dead and, as with the third film of the Godfather trilogy, another
composer had to be found. Morricone supplied a darkly coloured and eco-
nomical score, far removed from the Italianate lyricism of the director’s
earlier Shakespeare projects. The pulsating ostinati of the main titles sug-
gested the pagan mood of Carl Orff, with the use of organ for Elsinore
carrying a hint of the Gothic and the underscoring of the opening
internment of Hamlet’s father an echo of the solemnity of Wagner’s Parsifal.

The music was predominantly minor in key, with luminous major triads
reserved for the apparition of the ghost of Hamlet’s father and his son’s
swearing of revenge; elsewhere the tonality was designedly ambiguous, even
in the more expansive music underscoring dialogue between Hamlet and
Ophelia. As in Walton’s treatment of the same scene, the play-within-the-
play benefited from the blurring of diegetic and nondiegetic music in which
subtle use of percussion instruments played an important part. At no point
was there any attempt to sentimentalize or ennoble Shakespeare’s action (for
a contrasting treatment, see below), and only for the end credits did Morri-
cone compose a mournful oboe theme, bringing the drama to a conclusion
marked by dignified yet expressive restraint.
No box-office success remotely comparable to that of Zeffirelli’s Romeo
and Juliet attended Roman Polanski’s film of Macbeth (1971), the com-
mercial failure of which may well have contributed to the paucity of film
adaptations of Shakespeare in the 1970s and 1980s. (A further factor was
the saturation of the international television market by the ambitious series
of studio Shakespeare productions undertaken by the BBC at this time.)
Controversial at the time for its nudity and graphic violence – the climactic
demise of Macbeth at the hands of Macduff remains one of the grisliest
decapitation scenes in the history of cinema – the film featured idiosyn-
cratic music by The Third Ear Band. Their intimate quasi-theatrical score
deployed restrained avant-garde twitterings and scratchings for the irra-
tional elements of Shakespeare’s play – witches, apparitions, ghosts and
paranoid dreams (Macbeth’s ‘sorriest fancies’, as Lady Macbeth puts it) –
and simple lyricism (oboe and guitar in unison melody) for moments of
tenderness between Lady Macbeth and her husband. A remote Celtic atmo-
sphere was created by using folk-like drumming, bagpipe sonorities with
dissonant drones and designedly out-of-tune wind instruments. Very rarely
was background music called upon to create additional tension in the drama,
this being generally achieved through an absence of music altogether or, in
one memorable instance, agitated diegetic dance music underpinning Lady
Macbeth’s taunting of her husband with an accusation of cowardice. As
in Rózsa’s otherwise very different score for Julius Caesar, Polanski’s Mac-
beth featured simple ritualistic drumming to intensify the preparations for
impending battle.
Non-Anglophone cinema has – with the exception of the Italian films
discussed above – made relatively sparing use of Shakespeare’s plays in rec-
ognizable form. The foreign director who engaged most thoroughly with
Shakespearean scenarios was Akira Kurosawa, whose Throne of Blood (1957)
was based on Macbeth, and Ran (1985) was a Samurai reinterpretation of
King Lear. (For discussion of the music in Kurosawa’s films, see Chapter 9.)
Film-makers in the Soviet Union made several notable films of Shakespeare
in Russian translation, beginning with Sergei Yutkevich’s Othello in 1956,

with music by Aram Khachaturian that (in a time-honoured gesture) quoted
the Dies irae chant for the murder of Desdemona. Director Grigori Kozintsev
commissioned Shostakovich, who had worked with him many decades
before, to provide scores for his film versions of Hamlet (1964) and King
Lear (1970), both films using the Russian translations by Boris Pasternak.
Shostakovich had already composed music for theatrical productions of the
two plays, and enjoyed a close working relationship with Kozintsev, who
often gave the composer unusually precise instructions about the nature of
specific musical cues.
In Shostakovich’s music for the film Hamlet, parallels with the style of
the composer’s later symphonies are discernible, and Egorova has gone so
far as to call the score a ‘programme symphony’ for which she proposes a
sonata-form archetype for the organization of the themes for Hamlet, the
ghost of his father, and Ophelia (Egorova 1997, 172–3). While this interpre-
tation may be regarded as somewhat forced, it is nevertheless notable that
the style of Shostakovich’s film music is generally consistent with that of
his concert works and makes no concession to commercial considerations.
Three of the composer’s trademark idioms are prominent in Hamlet: acerbic
scherzo writing (for the strolling players), austere music of grim defiance,
and a brittle militarism. In addition, the composer includes naı̈vely styl-
ized harpsichord music for Ophelia’s dancing lesson and a lilting melody
in compound time, also for Ophelia, inhabiting a soundworld of archaic
lyricism; that Shostakovich intended to tap Old English associations is indi-
cated by his reuse of the same melodies for Ophelia’s songs as those in the
Walton–Olivier film, which had originated in productions of Shakespeare’s
play at London’s Theatre Royal, Drury Lane, in the eighteenth century and
were published in 1816 (Ford 2007, 24). A marvellously agitated climax
accompanies Claudius as he stumbles from the players’ dumb show in Koz-
intsev’s film, aware that they are portraying his heinous crime, and much
of Shostakovich’s score is saturated by variants of Hamlet’s theme. (The
realization that the latter bears a generic similarity with Shostakovich’s own
musical monogram, as used in the Tenth Symphony and Eighth String
Quartet, will intrigue those keen to trace autobiographical elements in the
composer’s work.) Kozintsev’s interpretation of the play reflects Soviet ide-
ology in his view of a fundamentally heroic Hamlet, and this is mirrored
in Shostakovich’s music: here the man of action overcomes his vacillations,
and his heroism is celebrated at the film’s end in forthright funeral music,
‘not an epitaph to a hero who perished in vain, but a baton handed over to
future fighters’ (Efim Dobin, quoted in Egorova 1997, 184).
Shostakovich’s score to King Lear again exemplifies three of the com-
poser’s most characteristic moods: bleakness, restrained wit and stark
militarism. In several places the music makes a contribution to increas-

ing dramatic tension, especially when Lear storms out (rather like Claudius
in Hamlet) after pairing off his rejected daughter Cordelia with the King of
France, and when the militaristic music for images of wild horses builds to
a climax – although the effect here is weakened by the abrupt lowering of
the music’s volume when it threatens to compete with dialogue. The most
memorable musical gesture in the film is the simplest: the gloomy melody
for an unaccompanied and thinly toned E flat clarinet that accompanies
the opening titles, strikingly economical and suggestive of music for a live
stage production, is later revealed to be the Fool’s theme when he performs
it diegetically on his pipe as Cordelia and Lear are reconciled. This takes an
effective liberty with the original text, in which the Doctor’s line ‘Louder the
music there!’ is not addressed to the Fool, who takes no part in the action
at this point. In the words of Egorova, the melody combines ‘sorrow with
a lucidity which leaves hope for purification through suffering’ (Egorova
1997, 221).
The most prominent purveyor of filmed Shakespeare in the last decade
of the twentieth century was British actor/director Kenneth Branagh, whose
Renaissance Films company drew on the success of stage productions by the
Renaissance Theatre Company. Composer Patrick Doyle joined Branagh’s
troupe in 1987, and went on to compose the music for all the director’s
Shakespeare films to date. Their cinematic collaboration began with Henry
V (1989), widely regarded as a conscious critique of Olivier’s interpretation
of the play. Although a period setting was retained, the cinematography was
more up-to-date and the Agincourt scenes in particular strove to achieve
a mud-and-gore realism more appropriate to the post-Vietnam era. Much
of Doyle’s music was concise and eloquent, with traces of the contrast-
ing influences of (spiky) Shostakovich and (noble) Elgar, and some boldly
rhetorical orchestral gestures contrasting with laconic simplicity (as when
the King of France is introduced by unaccompanied bass clarinet). Full
string textures based on ostinati were the perfect accompaniment to images
of the English doggedly marching through the mud, and the choral set-
ting of Non nobis (initiated diegetically by Doyle himself, playing the role
of Court) was, in spite of its rather naı̈ve optimism, a rousing conclusion
to the battle sequence – so much so that, at a screening in New York, the
audience mistakenly assumed that the hymn indicated the end of the film
and began to leave the theatre (Rosenthal 2000, 41). In several places, how-
ever, Doyle and Branagh included music cues without compelling reason,
and Doyle’s tendency to indulge in his neo-Elgarian nobilmente idiom as
a means of dignifying a scene – a technique which becomes increasingly
prominent in the second half of the film – at times seemed overdone. In
places, the uncomfortable disparity between image and music seems to have
resulted from a lack of insight rather than deliberate anempathy. As Samuel

Crowl has commented of Doyle’s scores in general, they are ‘emblematic of
Branagh’s desire to employ an extravagant film vocabulary’, and in Henry V
the music ‘romanticises the English victory in a way that the battle’s images
do not, opening the door for Branagh’s detractors to accuse such moments
in his films of being ideologically unstable and politically pernicious’ (R.
Jackson 2000, 228). One particular scene in the battle sequence combined
slow-motion photography of brutal fighting with incongruously expressive
music, an anempathy directly echoed in Ridley Scott’s Roman epic Gladiator
and other modern Hollywood epics (see Chapter 12).
The tendency to include musical cues at every obvious opportunity grew
more acute in Branagh’s Much Ado About Nothing (1993). The film’s undis-
criminating spotting has been criticized by one Shakespeare scholar for
using almost continuous music ‘to smoothe over the prickliness of the text.
By inflating the play’s brittle epitaph scene (Act 5 scene 3) with orchestra
and full chorus accompanying a procession of penitents, Branagh made
the sequence operatic and compounded the film’s leaning towards melo-
drama’ (Hattaway in R. Jackson 2000, 93). In the musing soliloquies spoken
by both Beatrice and Benedick as they overhear scheming gossip, Doyle’s
music fluctuates between superficial comedic gestures and a gratuitous sen-
timentality that weakens the liveliness of Benedick’s soliloquy in particular.
Such accusations hark directly back to the earliest critical anxieties about
the application of easy-listening background music to Shakespeare’s sophis-
ticated and often disquieting poetry. As in Henry V, the score featured songs
performed diegetically by Doyle himself, that sung at Hero’s tomb descend-
ing into cloying sentimentality. The main-title music, however, was both
epic and colourful, building in bustle and excitement to end on a well-
timed ceremonial flourish as the two opposing groups of men and women
finally came face to face.
Branagh’s Hamlet (1996) set the complete text of Shakespeare’s play in
a film lasting some four hours. The production, which located the action
in the nineteenth century, was distinguished in its acting and cinematogra-
phy, while Doyle’s score included felicitous moments such as the ceremonial
march taking us into the highly stylized interior of Elsinore, and the disqui-
eting string music for Ophelia’s flashbacks to her lovemaking with Hamlet.
Yet in three fundamental regards, all adumbrated in his earlier work, Doyle’s
score was disappointing. In the first place, music cues were so numerous
(and often unnecessarily prolonged) that they often failed to have much
impact – a problem made more acute by the extreme length of the film as a
whole. Second, neutral music with fussy detail frequently competed directly
with dialogue while adding nothing to the prevailing mood. Third, Doyle’s
quasi-Elgarian ‘ennobling’ music seemed directly at odds with the dramatic
4.3 Shakespeare meets the Hollywood musical headlong in Kenneth Branagh’s flamboyant Love’s Labour’s Lost
(1999). The director is second from right.
action in a number of important places, for example when Hamlet snarls

‘O most pernicious woman!’ (of his mother) and spits out a tirade against
his environment (‘the earth seems to me a sterile promontory, this most
excellent canopy the air . . . appeareth nothing to me but a foul and pestilent
congregation of vapours’), and when – in the most supremely undigni-
fied moment of the entire play – he wrestles on the ground with Ophelia’s
brother at her funeral. The use of such music to provide ‘depth’ to a scene
had, not for the first time in the history of film music, evidently become a
stock and clichéd response applied without due consideration being paid to
the relevant dramatic context.
In the late 1990s, both traditional and updated settings for filmed
Shakespeare remained equally viable. A new version of Othello (1995, dir.
Oliver Parker) with a sumptuously realistic period setting received a som-
bre and economical score by Charlie Mole. Far more flamboyant was Baz
Luhrmann’s William Shakespeare’s Romeo + Juliet (1996; music by Nellee
Hooper, Craig Armstrong and Marius de Vries), which was squarely aimed
at the youth market and followed the contemporary fashion for including
the name of the author of ‘classic’ texts in the film’s title – presumably so
that no unsuspecting cinema-goers could reasonably ask for their money
back afterwards. (Lest this comment should seem unduly cynical, it might
be noted that when a British national newspaper ran a multiple-choice

general-knowledge test for young people in the late 1980s, the majority of
participants stated that The Tempest was written by pop star Kylie Minogue.)
Luhrmann’s contemporary West Coast setting is infused with images of the
media, and the film’s visual style dominated by the dynamism and restless-
ness of MTV pop videos; the music track veers with equal unpredictabil-
ity from a sombre choral-orchestral style inhabiting territory somewhere
between Orff and Mozart’s Requiem (used for long-shot cityscapes), to
heavy rock beats and pop songs (both diegetic and nondiegetic). This was
not the first Shakespeare film rooted in contemporary popular culture –
Derek Jarman’s interpretation of The Tempest (1979) had a punk ambi-
ence, with electronic music by Wavemaker and a deliciously camp wedding
scene bringing together sailors dancing to the panpipes of Gheorge Zam-
fir and a bluesy rendering of ‘Stormy Weather’ by diva Elisabeth Welch –
but it was certainly the most exhilarating. In a quite unexpected shift of cul-
tural emphasis, Luhrmann imposed the high romanticism of the ‘Liebestod’
from Wagner’s Tristan und Isolde on the denouement of the tragedy. Clas-
sical music also features earlier in the film, when the pompous bustle of
the preparations for the ball at the Capulets’ mansion is accompanied by
Mozart. Shot through with self-parody and occasional parodies of other
genres (e.g. the initial confrontation between Montagues and Capulets at
the gas station, in which both visuals and music ape the conventions of
the spaghetti western), the film is sensitively spotted, and often makes its
musical points economically. Some techniques are conventional, such as the
use of diegetic music followed by the suppression of ambient sound when
Romeo and Juliet first meet; here, as in the sudden outburst of lively music
at the end of the first balcony scene, there are clear parallels with Zeffirelli’s
treatment of the same play.
Richard Loncraine’s film of Richard III (1996; music by Trevor Jones)
sets the violent action in a fictional 1930s civil war and charts the increasing
fascist domination perpetrated by Richard’s faction, with obvious parallels
to the rise of the Nazi party. The dark scoring, with its sultry saxophone
theme, fits comfortably into a modern thriller style, but more memo-
rable is the handling of popular music appropriate to the period setting.
A poem by Shakespeare’s contemporary Christopher Marlowe, ‘Come live
with me and be my love’, is given a full-blooded diegetic setting in the style
of big-band swing at the opening of the film, and this infectiously lively
music returns in the underscore when Richard coldly decides to take a wife,
neatly reflecting the capricious amorality of his Machiavellian manipula-
tions. Richard’s coronation is seen from Anne’s drugged perspective, with
Charpentier’s Te Deum distorted before the scene turns monochrome and
the recording quality of the soundtrack is suitably downgraded as we then
flash forwards to see Richard viewing the ceremony as a movie projection.

Novel use of a film-within-a-film also distinguished Michael Almereyda’s
updating of Hamlet (2000; music by Carter Burwell), in which the dumb
show is played out as a silent film accompanied by a medley of traditional
scoring styles.
Branagh’s flamboyant interpretation of Love’s Labour’s Lost (1999) drew
on sparkling arrangements of 1930s hit songs that form the basis for brief
but compelling dance numbers paying homage to the Hollywood musi-
cals contemporaneous with the updated action, set in Oxford around the
time of the Second World War; the narrative is clarified by newsreel-style
interludes throughout. Branagh’s comic conception works brilliantly, and
the film is a satisfying celebration of Hollywood classicism, canonic popu-
lar song and timeless poetry. In Michael Hoffmann’s lacklustre version of
William Shakespeare’s A Midsummer Night’s Dream (1999; music by Simon
Boswell), the updating of the action to a bicycle-infested nineteenth-century
Italy did not mask the fact that the soundtrack was deeply conventional, and
in some respects harked directly back to the earliest Shakespeare sound films.
Like Reinhardt’s 1935 treatment of the same play, Mendelssohn’s incidental
music was prominently featured: in the main-title sequence, Mendelssohn’s
Overture (at first crudely edited and then recomposed in pastiche style)
accompanied an animation of flitting fairy lights which would not have
been out of place in Disney’s Fantasia, while the Nocturne and Wedding
March appeared at relevant points in the plot. Predictably, the most colour-
ful scoring was reserved for the supernatural elements of the play, and
mickey-mousing employed for comedic moments, while the drinking song
from Verdi’s La traviata and other Italianate material served as locational
music.
In addition to the many films directly based on Shakespeare’s plays,
numerous others have taken the bard’s work as a starting point. Discus-
sion of the music for some of these films will be found elsewhere in the
present volume: see, for example, Forbidden Planet (a sci-fi reworking of
The Tempest), West Side Story (a musical based on Romeo and Juliet) and
the Greenaway–Nyman collaboration, Prospero’s Books (also inspired by
The Tempest). Perhaps the starkest illustration of the wide range of musical
styles acceptable when dealing with Shakespearean themes, even in Eliza-
bethan costume, are Tom Stoppard’s two comedy screenplays, Rosencrantz
and Guildenstern are Dead (dir. Stoppard, 1980; music by Stanley Myers) and
Shakespeare in Love (co-writer Marc Norman, dir. John Madden, 1998; music
by Stephen Warbeck). Rosencrantz, which was itself originally a stage play, is
a devastatingly witty gloss on Hamlet seen through the eyes of two of Shake-
speare’s shadowy minor characters, and the surrealism of the text is per-
fectly captured in the incongruous down-to-earth blues that accompanies
the front and end titles, while the dark undercurrents of the film are hinted at
by atmospheric electronics. In contrast, the Oscar-winning music for Shake-
speare in Love, like that film’s much less intellectually demanding screenplay,
is an easy-listening romantic score in keeping with the commercial bias of
the enterprise.
The last major Shakespeare film of the twentieth century was one of
the finest. Julie Taymor’s Titus (1999) was a visually stunning adaptation
of the bard’s grisly revenge tragedy, Titus Andronicus, its production design
a surreal mixture of ancient, 1930s and modern elements and its score
consistently resourceful. As a representative compendium of the manifold
scoring techniques available to the film composer at the turn of the cen-
tury, Elliot Goldenthal’s music for Titus is exemplary. The initial and highly
stylized approach of Titus’s army is accompanied by a thumping march
for percussion, which yields to austere choral incantations with orchestral
accompaniment in a style reminiscent (like Morricone’s Hamlet music) of
Orff’s Carmina Burana, here tinged with the swirling paganism of Mus-
sorgsky’s Night on Bald Mountain and setting a Latin translation of part
of Shakespeare’s text. For the first daylight exterior shot of a perplexingly
futuristic ancient Rome, the idiom switches abruptly into a ‘boogie-cool
jazz amalgam’ (Taymor 2000, 182) to accompany the entrance of Saturni-
nus’s motorized cavalcade. Thereafter the soundtrack presents a myriad of
stylistic allusions, including Stravinskyan fanfares, shades of John Adams’s
dynamic minimalism, a Mahlerian ‘fate’ motif (juxtaposing a high major
third with a low minor triad), an up-dated swing-band idiom with electric
guitars for the first orgy scene, an atonal walking bass line in free-jazz style
for the brawl between Tamora’s sons, manically pulsating hunting music, a
bizarre scherzo leading up to the lopping off of Titus’s hand, head-banging
electronics (produced by Richard Martinez) for the arcade games in the den
of Tamora’s murderous sons, and Carlo Buti’s sentimental old song ‘Vivere’
to accompany the serving up of the sons’ flesh in meat pies for unwitting
consumption by their mother. The jazz and popular elements reinforce the
black comedy of the play’s more grotesque scenes, as when Titus’s hand is
dropped neatly into a polythene bag for delivery to the Emperor, and kalei-
doscopic carnival music (of a kind Goldenthal previously explored with
Taymor in the theatre) accompanies the shocking moment when the hand
is returned to Titus along with the severed heads of his own sons. In spite of
such manifold eclecticism, the film’s strong visual style and first-rate acting
both lend a compelling sense of coherence to the whole, and Goldenthal’s
extraordinary musical journey culminates in a cathartic end-title fantasy of
considerable beauty.

4 - Stage and Screen

Uploaded by

Copyright:

Available Formats

4 - Stage and Screen

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

4 - Stage and Screen

Uploaded by

Copyright:

Available Formats

Cambridge Books Online

A History of Film Music

Book DOI: http://dx.doi.org/10.1017/CBO9780511814341

Online ISBN: 9780511814341

Hardback ISBN: 9780521811736

Paperback ISBN: 9780521010481

4 - Stage and screen pp. 131-182

Chapter DOI: http://dx.doi.org/10.1017/CBO9780511814341.005

Cambridge University Press

and relatively sophisticated in technique, and later operatic ventures were

It is worth reiterating, however, that the first screening of the Rosenkavalier

in a street scene (with the recording level manipulated to suggest distance in

cash settlement and potentially lucrative options to score further films by

several early commentators on film music bluntly predicted no future for

In his account of the problematic nature of filmed opera, London concluded:

the basic unreality of the events, which is inherent in the words, is

Significantly, Schoenberg was considering Roller as one of three possible

in the score between vernacular musical idioms and conventional operatic

Film in opera; opera in film

and ‘noumenal’ in opera: see Abbate 1991.) In Werner Herzog’s Fitzcarraldo

Herzog, Nicholas Hytner, Losey, Rouben Mamoulian, Ken Russell, Luchino

The film musical

The backstage musical acquired stunning visual interest and choreo-

fetishistic and elaborate chorine formations, Berkeley’s choreography and

everyday activities of an enclosed community. The new style was evident in

standard in rhythmicized vernacular song style directly inspired by physi-

constitute ‘reverential attempts, in a blockbuster era of Hollywood desper-

Astaire is the consummate theatrical dancer, while Kelly is more interested

De Mille believed that each gesture should be a natural extension of the

pointedly contrasted to the depressive and egotistical unemployed classical

While everything else is undergoing radical disruption, the music is at its

Street, a film distinguished not only by its perfect wedding of nightmarish

of Shakespeare’s filmed plays might otherwise be interpreted (Hammond

a rash of film productions based on non-copyright literature and opera,

in a manner recalling the great Berkeley musical routines, was dismissed

One ‘good modern composer’ who soon made an international reputa-

as when the voice-over describing Ophelia’s suicide is backed by a recitative

music by Rota, that did most to establish Zeffirelli’s international reputation,

internment of Hamlet’s father an echo of the solemnity of Wagner’s Parsifal.

in Russian translation, beginning with Sergei Yutkevich’s Othello in 1956,

militarism. In several places the music makes a contribution to increas-

resulted from a lack of insight rather than deliberate anempathy. As Samuel

action in a number of important places, for example when Hamlet snarls

be noted that when a British national newspaper ran a multiple-choice

flash forwards to see Richard viewing the ceremony as a movie projection.

You might also like