The murder of Grandi

The second example is the scene of Grandi’s murder (1958: 1h 9’ 29’’-1h 14’ 56’’; 1976: 1h 20’ 44’’-1h 26’ 10’’; 1998: 1h 16’ 38’’-1h 22’ 5’’) which takes place in the Ritz Hotel. During their investigations into the death of Linne­kar, Vargas and Quinlan clash over the line to be taken during interrogations; discovered by the Mexican as he was busy falsifying evidence to be used against a suspect, Quinlan decides to eliminate him by blackening the reputation of his wife. Suzan Vargas is seized by members of the Grandi family who stage an orgy, leaving the woman lying in a state of shock with marijuana and heroin scattered about the room. After making an anonymous phone call reporting the whereabouts of Su­zan, Quinlan kills Grandi in the room where she is asleep so as to place the blame for the murder on her and get rid of an awkward witness. All three ver­sions of the film are characterised by a precise atmosphere created, in sound terms, by a modification of the music. At the start of the scene the audio track sounds like music coming from a juke box or a band playing on the ground floor of the hotel: the audio has a muffled quality and the music, exemplifying Welles’s ‘background’ cate­gory, seems to be filtered through the walls of the building. The scene’s finale is also the same in all three versions: the music, turned up louder and with the filters removed, is used to comment on the ac­tions taking place with increasing intensity. The brass instruments and the conga rhythm provide a counterpoint to the sounds of the struggle between Grandi and Quinlan, with changes of rhythm marking formal breaks in the ac­tion. This new ‘underscor­ing’ function becomes ever more apparent as the scene progresses: Grandi’s desperate shouts are echoed by the trumpets and his la­boured breathing by the percussion, while Suzan’s cries on seeing the corpse are pre­ceded, introduced almost, by a trumpet blast. By contravening the laws of acoustics and preventing the spectator from identifying with a realistic lis­tening point – the band or juke box would have had to enter the room where the mur­der takes place – this stratagem endows the scene with a surreal, oneiric atmosphere that can clearly be referred to Suzan: still in a state of drowsiness, but shortly to become the scene’s principal viewpoint. There is, however, one detail that changes: in the 1958 and 1976 versions the passage of the music’s status from background to underscoring corresponds to Grandi’s smashing of the window during the struggle of the two men (1958: 1h 13’ 31’’; 1976: 1h 24’ 45’’). Whereas in the 1998 version the sound begins to increase in intensity when an exasperated Quinlan puts the phone down and locks the door (1998: 1h 19’ 55’’). The decibels increase, and the sound starts to lose the characteristic muffled quality of the early part of the scene, before the window is smashed. Thus in Schmidlin and Murch’s version the trans­formation in the musical element is not justified by a narrative element (on the contrary, closing the door should have produced the opposite effect, making the sound even more indistinct). The shattering of the window loses its structural function of separating off interior and exterior, real and surreal, and becomes a mere scenic event. The different chronology alters the effect of the scene: in the first two versions the music be­comes underscoring when Suzan begins to regain consciousness, anticipating the change in perspective for the ensuing narrative; in the third it becomes commentary slightly ear­lier.

One further clarification is called for: in strictly logical terms the example we have just illustrated has nothing to do with ‘the author’s wishes’, since the entire audio component, and hence also the progressive increase in decibels, was the result of the editing carried out by Universal in 1958. In this respect the work of Schmidlin and Murch’s team was arbitrary and surreptitious – since, unlike other interventions, they did not explicitly refer to this one – and merely based on the version of 1958. Indeed it may be thought that my choice of this example is methodologically questionable. It does however illustrate the risks inherent in acting according to ‘the author’s wishes’: Welles made no criticism of the scene in question, giving the impression of being satisfied with its rendering, or at least of not considering it disastrous. Thus it could presum­ably have featured in a ver­sion which met with his approval. As a matter of fact the transformation in the status of the musical element, which takes place ‘in direct’ in front of our very ears, as it were, almost as if to reveal the mechanism commonly used in Holly­wood sound tracks, stands as an original meta-textual reflection which the director could have evoked during the filming or which, more simply, he could have adopted at the suggestion of a collaborator.