The audience as creative contributor: examining the effects of implied sound and music in two versions of Nosferatu: A Symphony of Horror (1922)

ABSTRACT Since the release of Giorgio Moroder’s restoration of Metropolis in 1984, paired with his electronic musical score, it has become increasingly popular for composers to create new scores for silent films. Some scholars credit new musical scores with introducing younger generations to silent cinema. However, other critics argue that new scores make silent films less authentic. F.W. Murnau’s Nosferatu: A Symphony of Horror was initially released with an orchestral score by Hans Erdmann. This original score is still regularly used to accompany Murnau’s film when it is screened today and is contained on the Eureka Video home media releases of Nosferatu that feature the 2013 F.W. Murnau Stiftung restoration of the film. In 1997, composer James Bernard created a new score to accompany Photoplay Productions’ remaster of Enno Patalas’ 1995 Nosferatu restoration, which has since been paired with the BFI DVD and Blu-Ray releases of Nosferatu. This article examines the effects that Erdmann’s and Bernard’s musical accompaniments have on implied sound in Murnau’s film in the two aforementioned editions of Nosferatu. In particular, I focus on how audible and inaudible sound can affect the audience’s reception of, and creative engagement with, Nosferatu.

film (see Massaccesi 2016, 42;Audissino 2019, 181). Furthermore, although critics acknowledge that Nosferatu has a musical subtitle (A Symphony of Horror) (see Collier 1988, 149-150;Roberts 2008, 42-43;Massaccesi 2016, 91-92), its use of implied sound is rarely discussed. To be clear, by using the term implied sound, I am not referring to audible sound that is heard by audiences acoustically in the cinema. Instead, I term implied sound to be the internal manifestation of sound that is understood by spectators through visual and textual sound stimuli in silent films.
This article seeks to expand the current discussions of Nosferatu's music by Massaccesi and Audissino, in addition to the limited analysis of implied sound in Murnau's film, by specifically examining how musical accompaniments can affect our creative engagement with implied sound and determine our reception of Nosferatu's narrative and characterisations. The two versions of Nosferatu that I examine are the 2013 Eureka Video and 2015 BFI Blu-Ray editions of Murnau's film. The former features a 2013 restoration of Nosferatu completed by the F.W. Murnau Stiftung and is paired with Hans Erdmann's original 1922 score. Conversely, the latter contains the 1997 Photoplay Productions remaster of Enno Patalas' 1995 reconstruction of Nosferatu and is accompanied by James Bernard's 1997 score for the film. I have selected the Eureka Video and BFI variants over other Nosferatu versions because they are regularly screened in cinemas and are the most accessible editions of the film currently available on UK home media today. 1 I seek to demonstrate that, despite exclusively making silent films, Murnau had an acute understanding of sound and passion for music.
I argue that Erdmann's score, which was produced in partnership with Murnau (see Patalas 2002, 30), preserves the film's ambiguous implied sound by typically avoiding the use of musical sound effects. This approach enables the audience to shape their own interpretations of Nosferatu's narrative and characterisations because spectators are given the autonomy to decide exactly what noises exist in the film world and how they should sound. In contrast to the original score, Bernard regularly employs sound effects to audibly define implied sound throughout his 1997 accompaniment and music to remove ambiguity from the film's plot. I contend that this approach from Bernard eliminates Murnau's implied sound, and thus prevents the audience from creatively engaging with Nosferatu and constructing their own unique readings of the film, as they are forced to accept that specific noises exist and sound a particular way. Therefore, despite adding a new layer of audible sound effects to Nosferatu, we could interpret that Bernard's score actually makes Murnau's film less engaging and complex than the original silent version was when it was first released in 1922.
To help undertake my study of implied sound in Murnau's Nosferatu, I aim to answer a number of research questions. How is implied sound produced in Nosferatu and what is its function in the film's narrative? How does an examination and comparison of Erdmann's and Bernard's scores for Nosferatu change our understanding of Murnau's approach to implied sound in the film? To what extent and how do Erdmann's and Bernard's scores for Nosferatu affect the audience's reception of, and engagement with, Murnau's silent film in two different editions of the film?
In the first part of this article, I explore our current understanding of silent film accompaniment, the creative role that the audience can have during silent film exhibition, and the trend towards rescoring silent films in the post-silent era. Following this, I outline Murnau's relationship with music and his attitude towards sound as a silent film director. I then proceed to analyse and compare Erdmann's and Bernard's musical scores for Nosferatu, by examining the effects that they have on implied sound in key scenes in Murnau's film. In addition to this, I also articulate how the combination of implied sound and music in the two versions of the film can affect the audience's engagement with, and reception of, Nosferatu's narrative.

Silent film accompaniment, the audience, and rescoring silent cinema today
Over the course of the last thirty years, Richard Abel and Rick Altman note that it has become a 'cliché' (Abel and Altman 2001, xiii) to acknowledge that, despite the medium's name, silent films were in fact rarely mute. Marc Silberman argues that this is because silent films were routinely accompanied by various forms of sound during exhibition. 'We need to recognize that the early cinema was not silent in the sense of soundless; sound had always been present in cinema auditoriums' (Silberman 2010, 42). The most common form of sonic accompaniment for silent films was music played during exhibition (see Milsome 1979, 109;Prawer 1980, 29-30;Anderson 1987, 257). Musical accompaniment was often considered to be a vitally important element of the moviegoing experience during the silent era. For instance, Dorothy Richardson states that '[m]usic is essential. Without it the film is a moving photograph and the audience mere onlookers. Without music there is neither light nor colour' (Richardson 1927, 161). Silent filmmakers and critics both agreed that silent film scores had the power to influence the audience's understanding of a film's narrative (see Carli 1995, 318). Although audible sound effects were present in scores produced during the silent era, including Erdmann's original accompaniment for Nosferatu, they were usually only used for loud prominent sounds, such as gunshots or thunderstorms (see Milsome 1979, 30;Rabinovitz 2001, 176;Chion 2009, 7). This conservative approach to sound effects by most silent era accompanists like Erdmann, allows spectators to become creatively involved during exhibition and determine their own readings of characters and film plots by imagining the implied sound present in a silent film. When the sound era arrived in the late 1920s and early 1930s, films quickly embraced audible sound effects and dialogue and, to an extent, the important role that musical accompaniment had during the exhibition of silent films was forgotten. Leonid Sabaneev, for example, wrote in the mid-1930s that 'it had been possible to play anything' (Sabaneev 1935, 43) to accompany silent films. I disagree with Sabaneev's statement and argue that this cavalier attitude has led some post-silent era musical compositions, such as Bernard's score for Nosferatu, to replace implied sound with audible sound effects and thus reduce the audience's creative engagement with silent cinema plots and characters.
Kevin Brownlow and John Kobal argue that silent film audiences use their imaginations to perceive inaudible sound within a film's visuals.
Talk to people who saw films for the first time when they were silent, and they will tell you the experience was magic. The silent film, with music, had extraordinary power to draw an audience into the story, and an equally potent capacity to make their imaginations work. They had to supply the voices and the sound effects [. . .]. The audience was the final creative contributor to the process of making a film (Brownlow and Kobal 1979, 7).
Audiences are able to do this because they arrive at the cinema with sonic knowledge from everyday life that they can draw on during exhibition (see Chion 1990: 145;Marks 2000: xiii). Rudolf Arnheim explains this phenomenon in detail.
No one who went unprejudiced to watch a silent film missed the noises which would have been heard if the same events had been taking place in real life. No one misses the sound of walking feet, nor the rustling of leaves, nor the ticking of a clock. The lack of such sounds (speech, of course, is also one of them) was hardly ever apparent (Arnheim 1957, 33).
Audiences meet silent films halfway and examine the visuals and text that they are presented with in order to fill in the missing sound gaps and better understand plot points and character motivations. Jean-Paul Sartre gives a detailed account of the connection between silent film scores and the audience's reception of implied sound.
Above all, I liked the incurable muteness of my heroes. But no, they were not mute. Since they knew how to make themselves understood. We communicated by means of music; it was the sound of their inner life. [. . . .] I would read the conversations, but I heard hope and bitterness; I would perceive by ear the proud grief that remains silent (Sartre, cited in Anderson 1987, 261). This testimony from Sartre and the earlier arguments from critics show us that silent film musical accompaniment was indispensable for silent era film audiences and directly aided their creative participation during exhibition. Erdmann's original score for Nosferatu, which is included on the Eureka Video home media releases of Murnau's film, supports the audience's engagement with the film's narrative by preserving Murnau's equivocal implied sound. However, the creative potential of silent cinema came to an end with the rise of the sound film, largely due to the introduction of sound effects and synchronised dialogue, which create unambiguous audible sound (see Duras, cited in Chion 1999, 9;Coates 1991, 20). In connection to this, Bernard's 1997 score for Nosferatu disrupts the audience's active participation with Nosferatu because he uses sound effects, amongst other sonic techniques, to explicitly define the film's implied sound.
In the 1970s, a new interest in silent cinema discussion, exhibition (see Cooke 2010, 4), restoration, and accompaniment began to develop around the world. To a large extent, this is due to the beginning of the New Film History movement in film studies which systematically rethought and reconceptualised our understanding of early cinema history. Furthermore, the 1978 Brighton Congress of the International Federation of Film Archives (FIAF) gave attendees unprecedented access to hundreds of rediscovered silent films and was instrumental in forming lasting partnerships between silent film academics and archivists from around the world. Curiosity for silent cinema grew further in the 1980s, thanks in part to the founding of Pordenone Silent Film Festival and the release of Giorgio Moroder's 1984 restoration of Fritz Lang's Metropolis (1927), featuring a new electronic score by Moroder. Moreover, the 1990s and 2000s saw interest in silent cinema increase further with the release of silent film restorations on home media (VHS, DVD and Blu-Ray) and thanks to the widespread availability of silent films on internet video streaming services like YouTube. These media platforms have made silent film more accessible than it has ever been, and have enabled new generations of composers to forge substantial careers for themselves in the creation of countless and varied new musical scores for restorations or revivals of silent films. The revitalisation of silent cinema has created debate surrounding whether a silent film should be presented with its original musical score (if it survives) in order to preserve the authentic silent era experience (see Carli 1995, 306) or if a modern one should be constructed to try and interest a new generation in silent cinema (see Audissino, 2016, 56;Brown 2019, 188). However, I attempt to move beyond the ethical and practical debate and instead focus on how a new film score can change the way the audience interacts with a silent film. Although I criticise Bernard's 1997 score for reducing Nosferatu's narrative complexity, I do not claim that all silent film scores composed since the end of the silent era undermine the audience's experience. On the contrary, newly composed scores for restorations of silent films have the potential to add new meaning to them and attract younger generations to silent cinema. However, if new musical accompaniments regularly replace implied sound with audible sound effects, the silent films they are paired with will likely lose some of their complexity because spectators will no longer be able to creatively interact with them during exhibition.

Murnau, music and Nosferatu
Unlike his fellow silent film director Fritz Lang, who described himself as 'a musical ignoramus' (McGilligan 1997, 16), Murnau had a strong and lasting interest in music throughout his life. Kevin Jackson notes that Murnau even entertained the prospect of becoming a composer before pursuing filmmaking and argues that his silent visuals have a striking musical quality.
In one period of his youth, [Murnau] had expressed the ambition to become a composer, in the vein of Mahler; he never achieved this goal directly, but it is fair to think of the flow of images in film as analogous to the flow of notes and phrases and melodies in music, Murnau was a skilled composer in celluloid. Hence no doubt, the subtitle: 'A Symphony of Terror' (Jackson 2013, 69).
Murnau's passion for music naturally transitioned into his approach to silent filmmaking. Jo Leslie Collier argues that, despite being technically mute, Murnau's silent films have a musical quality.
Music was another important aspect of Murnau's mise en scene. This statement might at first seem a paradox since Murnau was a filmmaker in cinema's silent era. However, films in Germany (and elsewhere) regularly had scores written for them [. . .]. More importantly, Murnau knew as Reinhardt had known that a performance could (and even should) be musical whether or not it contained music (Collier 1988, 149).
The consistent musical feel of Murnau's work becomes more understandable when we recognise that he was directly involved with the creation of the scores for his silent films. One of Murnau's Hollywood assistant directors, Frank Hansen, claims, in Lotte H. Eisner, that the director would 'suggest the main elements of the score to the composer' (Hansen, in Eisner 1973a, 86). Patalas backs up this point with direct reference to Nosferatu's composer Erdmann, and Giuseppe Becce, who composed music for Murnau's silent films Der Letzte Mann (The Last Laugh, 1924) and Herr Tartüff (Tartuffe, 1925). 'Erdman and Becce have reported that, before completing his films, Murnau used to discuss the music with the composers' (Patalas 2002, 30). This evidence from Eisner and Patalas shows us that Murnau was intimately involved in the creation of the musical accompaniment for Nosferatu with Erdmann. Although it is unclear if Murnau was aware of Nosferatu's implied sound when he worked with Erdmann on the film's 1922 score, I contend that the director's dual control of the film's visuals, which contain Nosferatu's implied sound, and its musical score allows these two parts of his film to coalesce without the music dismantling the film's implied sound and the audience's creative engagement during exhibition. Conversely, Bernard's sole role as a composer sees him neglect Nosferatu's visuals and create a score that removes Murnau's implied sound and the audience's capacity to be creatively involved in Nosferatu's plot and characterisations.

Comparison of Erdmann's and Bernard's scores for Nosferatu in the respective Eureka video and BFI editions of the film
To summarise, Nosferatu is set in the fictional German city of Wisborg in the year 1838. Hutter (Gustav von Wangenheim), an estate agent, is tasked by his employer Knock (Alexander Granach) to leave Wisborg and journey east to Count Orlok's (Max Schreck) castle in the Carpathian Mountains, in order to offer him the abandoned building that sits opposite Hutter and his wife Ellen's (Greta Schröder) house. When Hutter reaches the Carpathian Mountains he breaks his journey at an inn. During his stay, Hutter causes great distress to the superstitious locals when he states that he will be visiting Count Orlok's castle. Hutter soon arrives at Orlok's castle and meets the decrepit and monstrous Count. Orlok quickly becomes sexually interested in Ellen when he sees her portrait and creepily tells Hutter 'your wife has a lovely throat'. Meanwhile, Ellen seems to develop a mysterious illness that manifests itself at night and motivates her to sleepwalk and use telepathic communication. After being repeatedly attacked by Orlok, Hutter eventually realises that his host is a vampire. Thankfully for Hutter, Ellen saves her hapless husband by using telepathy. This act motivates Nosferatu (Max Schreck) to sail to Wisborg to claim Ellen, while Hutter races the vampire to his wife on land. Although Hutter arrives in Wisborg before Nosferatu, the vampire spreads bubonic plague through the city. After reading an occult book on vampirism, Ellen summons Nosferatu to her room and sacrifices herself, in order to kill the vampire and save Wisborg's remaining citizens.
One of the most significant ways that Erdmann's score actively supports the audience's creative involvement with the Eureka Video version of Nosferatu is through the use of mood music. This technique gently immerses the audience in the film's general atmosphere without determining their overall understanding of the narrative and characterisations. Erdmann's decision to habitually avoid sound effects allows Nosferatu's visual implied sound to remain ambiguous, which grants the audience significant creative freedom to interpret implied sound and construct their own unique readings of Murnau's film. For example, when Hutter arrives in the Carpathian Mountains tavern, Erdmann's mixture of soft and tense mood music illustrates the constant fluctuations in character emotions in the scene and helps the audience interact with implied sound. When Hutter enters the inn and excitedly sits down for his meal, Erdmann uses soft aerophone music to signify Hutter's and the patrons' serenity. However, when Hutter announces that he will soon be travelling to Count Orlok's castle, the aerophone music is immediately replaced with eerie string music. This abrupt change in music appears to be motivated by the shock felt by the inn patrons due to Hutter's destination and the subsequent unease that Hutter feels after witnessing the reaction of the local people gathered there. This use of music by Erdmann fills the audience with dread and informs them that characters' emotions have become tense, but does not seek to explain why the inn patrons feel this way about Count Orlok and Transylvania. Despite Murnau showing the tavern guests having animated discussions after Hutter reveals his intention to meet Orlok at his castle, Erdmann does not use sound effects to draw attention to the conversations or hint at their substance through particular musical sound effects. Do the people in the tavern know that Count Orlok is a vampire or are they just superstitious? Are they discussing Orlok or just talking amongst themselves about unrelated matters? It is left to the audience to study the visuals, music and the implied sound in the scene, and then come to their own conclusions on what the people are discussing and whether Hutter overhears their illuminating or innocuous conversations. In the BFI edition of Nosferatu, Bernard's use of music in this scene follows Erdmann's approach initially because he also displays the dramatic change of character mood through music when Hutter announces he is going to Orlok's castle. However, Bernard's score departs from Erdmann's towards the end of the scene when the innkeeper approaches Hutter and urges him to stay the night and avoid the werewolves outside, while other inn patrons huddle together and talk in the background of the shots. Throughout the duration of these conversations, Bernard employs fast-paced eerie musical notes that are synchronised with the mouth movements of the innkeeper and the other people gathered in the pub. In contrast to Erdmann, who elects not to draw attention to, or define, these conversations, Bernard chooses to make the audience aware of them and sonically suggest, through swift and eerie music, that they are discussing Nosferatu or something sinister relating to him. Consequently, and unlike Erdmann's accompaniment for this scene, Bernard's audience lose the ambiguity that is generated by Murnau's implied sound because Bernard makes the inn keeper's and patrons' voices, in addition to the substance of their conversations, clear through musical sound effects.
In the Eureka Video release of Nosferatu, Erdmann again uses unsettling mood music in a later tavern scene to help the audience understand that the characters feel uneasy. However, he does not use sound effects to replace implied sound and thus encourage the audience to construct a definitive reading. In this particular scene, Hutter has retired to his bedroom and looks out of the window at the horses that roam the grass outside. At this point, Murnau cuts to the forest near the field and focuses on a howling hyena, which is supposed to represent a werewolf, and introduces further eerie string music. Murnau then cuts to a group of old women, who cling to each other in terror inside the tavern. Due to Dracula's ability to turn into a wolf in Bram Stoker's 1897 novel, we can safely assume that we are supposed to infer that the old women in Nosferatu fear that the creature is a vampire. Following this, we again cut back to Hutter who closes his window and rubs his sides as if to suggest that he is cold. Unlike the old women who recoil in terror at the sight and sound of the hyena, Hutter does not appear to react in any way to the presence of the animal, despite looking outside and being in earshot. Although Erdmann's eerie music broadly signifies that the characters are unsettled in this scene, it does not attempt to guide the audience to a definitive conclusion by making implied sound audible. Instead, the audience is able to create the scene's visual implied sound and use this to help them self-determine the ambiguity surrounding the scene. For example, due to the fact that Hutter does not react to the sight or sound of the wolf, it raises the possibility that the animal is merely a figment of the old women's superstitious minds. Alternatively, perhaps the howling wolf does exist and Hutter feels undisturbed by it because he does not harbour the same mystical fear of these creatures? In any case, as with the other tavern scene that I discussed, it is up to the audience to piece together Murnau's visuals and Erdmann's music to create their own answers. Conversely, in the BFI edition of Nosferatu, Bernard's use of sound and musical effects in this scene removes Murnau's implied sound ambiguity and thus prevents the audience from creatively participating in the film. Although Bernard initially follows Erdmann and uses unsettling mood music in this scene, he changes course by using a sharp aerophone sound effect to audibly signify the wolf's screeching. As a consequence of this, the audience are not only prevented from internally crafting their own impression of the wolf's cry, through implied sound, but also forced to accept that the wolf is likely to be real because it creates audible sound. This invalidates one of the potential readings of this scene that I identified in Erdmann's score. Due to Erdmann's musical restraint, the audience could come to the conclusion that the wolf is merely imagined by the women in the tavern, because Hutter does not react to the sight or sound it makes. However, Bernard's accompaniment compels the audience to definitively accept that the wolf is real and Hutter is merely undisturbed by it.
In the scene when Nosferatu's carriage transports Hutter to the former's Transylvanian castle, Erdmann's and Bernard's contrasting approaches to implied and audible sound significantly affect the audience's ability to creatively interact with Murnau's narrative and characterisations. Shortly after Hutter has been abandoned by the coachmen that were initially tasked with transporting him to Count Orlok's castle and he begins the last portion of his journey on foot, Murnau inserts a shot of a fastpaced coach, which is driven by Nosferatu in disguise, jerkily moving through the countryside. The carriage appears again later in the scene and quickly speeds towards Hutter in a forest clearing, before swiftly taking him to Count Orlok's castle. Scholars have noted that Nosferatu's cinematographers were able to modify the pace and rhythm of the coach by under-cranking the camera in the shots that feature the carriage in motion (see Eisner 1973b, 105;Barlow 1982, 85-86;Calhoon 2001, 157;Kaes 2009, 103-105). The carriage's unnatural movement encourages us to view it as a supernatural vehicle. This is because the spasmodic and fast movements of Nosferatu's coach are strikingly different to the rhythm and speed of the other carriages in Murnau's film and the way that similar contraptions would move in real life. Throughout Erdmann's accompaniment of this scene in the Eureka Video edition of Nosferatu, implied sound is left intact because the composer does not use any sound effects for the carriage. Instead, Erdmann employs eerie string mood music for the duration of the scene, which sonically reinforces the mystical feel of Murnau's shots of the coach. The presence of implied sound in this scene allows the audience to creatively engage with the carriage by imagining what the magical contraption should sound like. Thanks to the lack of sound effects in Erdmann's score, the audience are not forced to accept that the coach sounds like a similar vehicle would in real life. On the contrary, as spectators, we have the freedom to study Murnau's paranormal visuals, in addition to Erdmann's eerie mood music, and, through implied sound, decide how the magical coach should sound. As the supernatural is not part of our reality and thus defies explanation and understanding, the presence of ambiguous and subjective implied sound in the Eureka Video version of Nosferatu reinforces the paranormal qualities of Murnau's camerawork and Erdmann's music by helping us to comprehend the carriage as a mystical object. The scene's implied sound does not seek to decode the mystery surrounding the vampire's coach. Instead, its equivocal quality actually contributes to the otherness of the vehicle because it is not granted any specific audible sounds that could make it appear more realistic to us. Conversely, throughout the course of Bernard's accompaniment of this scene, he employs musical sound effects that are reminiscent of how we would expect the coach to sound if we heard it in real life. In all of the shots that depict the Nosferatu's coach in motion, Bernard removes Murnau's implied sound by synchronising fast-paced drums and strings with the unnatural speed and rhythm of the carriage. These two musical sound effects evoke the recognisable real life sounds of horses' hooves repeatedly hitting the ground and wooden carriage wheels turning quickly. Not only does Bernard's use of sound effects prevent the audience from engaging with implied sound in this scene by deciding what the coach should sound like, it also throws the overall scene into disarray. Murnau's visuals suggest that the vehicle is supernatural, through the carriage's spasmodic and fast movement, but Bernard's use of drum and string sound effects sonically tie the contraption to the real world. Consequently, due to Bernard's elimination of implied sound and use of audible sound effects, we are motivated to read the coach as a semi-supernatural vehicle that is visually magical and yet sonically familiar.
In addition to Bernard's use of audible sound effects in his Nosferatu score, the composer also employs other musical techniques associated with sound cinema, particularly his Hammer films, which could affect the audience's creative engagement with Nosferatu's implied sound. For instance, in his Dracula (1958) score, Bernard frequently uses a leitmotif to suggest the word Dracula (Dra-cu-la) in scenes that feature the titular vampire. As Massaccesi and Audissino note, this technique was carried over into Bernard's score for Nosferatu, in which he regularly uses music to create the word Nosferatu (Nos-fer-atu) (see Massaccesi 2016, 42;Audissino 2019, 180). As a result of this, the audience's creative engagement with Murnau's silent film is often stunted or prevented entirely because it dismantles Nosferatu's implied sound and narrative ambiguity. An example of Bernard's use of the musical effect Nos-fer-atu, which eliminates the audience's creative interaction with Nosferatu's implied sound and narrative, occurs in a scene involving Knock. After Hutter has begun his journey to Transylvania, Knock is committed to the town's insane asylum for reasons unknown to the audience. Although Knock exhibited peculiar mannerisms before being confined to Wisborg's asylum, his behaviour appears to have become even more disturbing, as we see him eating flies in his cell while shouting 'blood is life' repeatedly. Throughout Erdmann's original score for Nosferatu, including in this scene, it is ambiguous as to whether Knock is a vampire or merely serving Nosferatu as his mentally-deranged human servant. Knock seems to desire blood for sustenance, as we see in the asylum scene, but at the same time, we never see him feed on a human or sleep in a coffin, which are two key components for Nosferatu's survival. Therefore, when viewing the asylum scene with Erdmann's score, the audience has the freedom to interpret Knock's implied voice as being suggestive of mental illness or monstrous to infer vampirism. In contrast to this, Bernard's musical accompaniment for this scene motivates the audience to accept a fixed reading of Knock's implied sound and character. Bernard's repeated use of the leitmotif Nos-fer-atu when Knock says 'blood is life' definitively paints Knock as a vampire, and eliminates the possibility that the character is mentally ill. As a consequence of this, the ambiguity present in Erdmann's reconstructed score for this scene is lost in Bernard's BFI accompaniment and the audience are encouraged to definitively accept that Knock's implied sound is demonic rather than maniacal.
Although Erdmann uses very few sound effects in comparison to Bernard's score for Nosferatu, it is important to address those that do occur and the impact that they have on implied sound and the audience's creative participation in the Eureka Video version of the film. For the most part, with the exception of the drummer herald and cockerel in Wisborg, Erdmann only uses musical sound effects for scenes that involve Nosferatu in Transylvania or Wisborg. The most concentrated use of sound effects in Erdmann's score are in the scene when Nosferatu approaches a defenceless Hutter in his castle bedroom. To begin with, we hear idiophone music which replicates the implied sound of the chiming clock in the castle. This causes Hutter to feel uneasy as he has been reading the book on vampires he obtained at the inn and realises that they feed at night. Hutter then opens the door and sees Orlok standing down the corridor staring at him. Following this, Hutter quickly closes the door again and hides in bed in terror as he realises that his host is a vampire. The next sound effect occurs when the door of Hutter's bedroom magically swings open, despite Nosferatu being far away from it. Erdmann uses a loud percussion sound to audibly indicate the sound of the door opening. After this, Erdmann once again uses idiophone music, this time for Nosferatu's slow and stiff steps towards Hutter's room. The use of sound effects comes to an end as Nosferatu is about to bite Hutter, but is repelled by Ellen's telepathic powers in Wisborg. Although it may initially seem that Erdmann's use of sound effects in this scene diminishes the audience's creative engagement with Nosferatu, it actually expands their understanding of the narrative and characters. It shows us that, unlike most other characters in Nosferatu, the dark magic that surrounds the vampire gives him extraordinary power. Nosferatu can not only artificially take control of people and objects visually, such as the castle door, but also manipulate and harness sound and communication too. The vampire is the only character in Erdmann's score for Nosferatu that can routinely create audible sound, while the rest can only imply it. Therefore, this scene causes the audience to view Nosferatu's sonic potential in an entirely different manner to the other characters'. Furthermore, it also encourages them to creatively engage with the scenes that involve Nosferatu communicating via telepathy, which are not accompanied with sound effects by Erdmann. The audience have witnessed the significant visual and sonic powers of Nosferatu in a physical environment. Therefore, when the vampire is locked in silent internal communication with characters, particularly Ellen, the audience has a platform to construct the implied sound of these telepathic conversations and add extra depth to their reading of Murnau's film. In contrast to Erdmann's use of sound effects, which give Nosferatu's plot and characters a new dimension, in addition to promoting audience engagement via implied sound, Bernard's employment of sound effects remove complexity from Nosferatu as they prevent the audience from creatively interacting with the film's implied sound.

Conclusion
In conclusion, through a discussion of implied and audible sound in Erdmann's and Bernard's scores for Nosferatu, this article has attempted to demonstrate that musical accompaniments can have a profound impact on the audience's reception of silent film plots and characters. Massaccesi and Audissino both recognise that Erdmann's and Bernard's scores are capable of influencing our understanding of Murnau's film, but neither analyse how the composers' different approaches to implied and audible sound affect spectators' ability to creatively interact with Nosferatu. For the most part, Erdmann's original 1922 score, which was produced in partnership with Murnau, consists of mood music and avoids audible sound effects. Consequently, when the accompaniment is paired with the Eureka Video edition of Nosferatu, Murnau's ambiguous implied sound is typically preserved and the audience have the autonomy to decide which noises exist in the film world and how they should sound. Although Erdmann's music attempts to immerse spectators in the eerie, romantic and tense moods that are present throughout Murnau's film, it does not try to dictate the audience's overall understanding of Nosferatu's narrative or characters. On the contrary, Erdmann's preservation of implied sound enables the audience to be creative participants during exhibition because they are not forced to accept definitive sonic readings of the film's characters and plot points. Murnau's dual control of Nosferatu's visuals and music ensures that neither element dominates the other or guides the audience's reception of the film. They both have important roles to play in the Eureka Video version of Nosferatu, but Murnau also leaves room for his spectators to shape their cinematic experience through implied sound. Conversely, thanks to Bernard's frequent use of audible sound effects in his 1997 Nosferatu score, implied sound is significantly reduced in the BFI edition of the film. By habitually replacing implied sound with sound effects, the audience are not able to internally create their own conceptions of the noises that are present in Murnau's film. In addition to sound effects, Bernard also uses a sonic technique from Hammer's Dracula that sees musical notes suggest the word Nos-fer-atu to the audience. Although this form of audible sound does not directly replace implied sound in the BFI version of Nosferatu, it nevertheless encourages us to interpret the phenomenon in specific ways that limit our reception of Murnau's plot points and characters. Unlike Erdmann's original score for Nosferatu, Bernard's accompaniment prevents spectators from creatively interacting with Murnau's film, through implied sound, which forces them to form narrow readings of Nosferatu. Note 1. Erdmann's accompaniment was thought to be lost for decades. However, his score was reconstructed in 1995 by musicologist Gillian B. Anderson and again by composer Berndt Heller in 2007 for the F.W. Murnau Stiftung's restoration of Murnau's film, which was released on DVD in 2007 and Blu-Ray in 2013 (see Audissino 2019, 176-177). Bernard is best known for his numerous musical scores for Hammer Horror films such as Dracula (1958). The 1997 BFI edition of Nosferatu was initially exhibited in cinemas and then released on DVD in 2002 and Blu-Ray in 2015.

Disclosure statement
No potential conflict of interest was reported by the author(s).

Funding
This work was supported by the Arts and Humanities Research Council (grant number AH/ L503848/1)