Chapter 2: The Elephant Man and Dune
The Auteur in Hollywood
Between Eraserhead and Blue Velvet, Lynch flirted briefly with the studio system. His relationship with Hollywood was initially fruitful, producers were keen to court the hot new director fresh off the back of his cult success. The first child of the marriage was The Elephant Man, a critical and commercial success (it received 8 Oscar nominations). It wasn't to last though, the relationship deteriorated and ended bitterly following the second production, Dune. Lynch missed his artistic freedom, and he was unhappy without the right of final cut, he said in 'Time' magazine, "I sold out on Dune". In the studio's eyes, Lynch had failed to live up to his commercial promise and they blamed him for delivering a 50 million dollar flop.
How this Auteur operated within the studio system raises a lot of issues around authorship. Andrew Sarris was aware of a distinction between "a 'commercial' director like Hitchcock and a 'pure' director like Bresson". Lynch had established himself with a 'pure' film and for the early part of the eighties he became a 'commercial' director. To what extent did the other elements of production influence the finished film, not only the producer but the co-scriptwriters, the original source material (both were based on books) and the actors involved. Is the Auteur compatible with the Star system. And, in answering the structuralist critics of the auteur theory, is Lynch's vision effected by the constraints of the genres he worked in.
The Elephant Man was based on two books, 'The Elephant Man and Other Reminiscences' by Frederick Treves (who features in the film both as a character and a cameo appearance), and "The Elephant Man: A Study in Human Dignity" by Ashley Montagu. The script was re-written by Lynch from an original draft by Christopher de Vore and Eric Bergren. The story is set towards the end of the nineteenth century and so fits the conventions of costume drama. The hero is John Merrick, the extremely deformed Elephant man of the title. It opens, like Eraserhead and The Grandmother before it with a stylised conception/birth sequence; a still photograph of his mother, a stampede of elephants, a woman writhing, her screams dubbed over with the sounds of elephants, a fade a burst of steam and a baby's cries. At a fairground, the doctor Treves steals through a door marked 'No Entry' to investigate a freakshow. Here he is shown Merrick by Bytes, Merrick's 'keeper'. Treves motives for his seeking the freak are left ambiguous, as is much of his character during the film. We are not sure how much is morbid curiosity, genuine concern or a desire for professional advancement (just as we don't know if Jeffrey is a detective or a pervert in Blue Velvet). Bytes points out to him how much they have in common when he pays for Merrick to be taken to the hospital and displayed for his academic colleagues.
Treves learns that Merrick has been beaten by Bytes and so houses him at the hospital. The hospital's administrator, Carr Gomm, is sympathetic but unsure about admitting an incurable. It is half an hour into the film, when a staff nurse is scared by the sight of Merrick, that the extent of the deformity is finally revealed to the spectator. The hospital's night porter becomes aware of Merrick's residency and exploits it for profit with arranged nocturnal viewing. As the hospital rehabilitates him by day, encouraging him to speak and introducing him into polite society, by night Merrick suffers increasing humiliation and exploitation, a paralleled double life. Bytes demands him back and inevitably, via one the night porter's clandestine visits abducts him. He is displayed at a fair in Ostend but is freed by his fellow freaks. On his return to London he is ill. At the station he is harassed by a group of children and chased by a crowd who corner him in a public toilet and unmask him. He screams to his tormentors, "I am not an animal, I am a human being." Treves recovers Merrick. They attend a pantomime together which the actress Merrick befriended dedicates to him. On his return to the hospital he opts to sleep lying down for the first time, a position he knows will kill him. The final image is the steam receding and his mother's eyes superimposed over a starscape with the litany "Nothing ever dies".
Dune was based on the Sci-fi novel by Frank Herbert. It is set in an unspecified future around a planetary system far from our own. The first forty-five minutes of the film attempts to pack in all the socio-political detail of the feud between the houses of Atredies and Harkonnen, and their battle to control Arakkis, the planet that is the sole source of melange, a spice that makes intergalactic travel possible. The story's focus is Paul Atredies, who is destined to be the Kwisach Saderach, a messiah. After 45 minutes the action moves to Arrakis, also known as Dune. House Atredies are taking over control of the planet from the Harkonnens. The Harkonnens have left many booby traps. Paul suspects the power on Arrakis lies not with the ruling house but with the Freman, the desert dwelling nomads of the planet. When the Harkonnens defeat the house Atredies, Paul and his mother are driven into the desert where they join the Freman, Paul becomes their messiah and riding the back of giant sand-worms they defeat the Harkonnens and the Emperor of the Universe in the process.
What differentiates the two films, and what was also probably a factor in one's success and the other's failure, is the amount of deviation the scripts took from their original source material. The Elephant Man takes great liberty with historical fact and constructs a story very similar in structure to Eraserhead; the opening birth sequence, the "sin" of Treves (the door marked 'no entry') as the catalyst to events that eventually culminate in the hero's death. The Elephant Man is probably the most accessible film lynch has made to date, and it is described by Michel Chion as "a film mercifully devoid of auteurial effects". I cannot agree with this as the film adheres just as closely to Lynchian convention as it does to the conventions of Hollywood narrative. The parallel worlds exist in the comparison between Merrick's diurnal and nocturnal existence. The Biological element is obvious, as are the copious theatrical references. Many of the other specifics have been detailed in the previous chapter.
Dune was a considerably more ambitious project. The budget was 50 million dollars. The book was a best-seller, therefore the story carried the expectations of a million fans' differing visualisations of characters and events. It also had a history of failed film attempts, with Alejandro Jodorowsky and Ridley Scott having previous plans to direct. Lynch, in this case, stuck rigorously to the novel, in fact probably too rigorously as he tried to pack in as much of the long and complex novel as he could. It resulted in what the critic Roger Ebert described as "one of the most confusing screenplays of all time".
With both films, there existed a script in some form before Lynch was hired as director. In both cases Lynch made rewrites to the script to bring it into conformation with his personal vision. In Dune's case the original script (by de Vore and Bergren, Elephant Man's co-writers) was totally rejected. Robert Lapsley and Michael Westlake, in "Film Theory: an introduction", are stating Cahiers du Cinema's opinion when they say "by taking a script written by someone else and by imposing his directorial style, an auteur makes the film his own". This is believable with Elephant Man but with Dune, the weight and complexity of Herbert's world didn't leave much room for Lynch's. Lynch had reputedly never heard about the book when first approached to direct the film. Herbert's thematic content is unfamiliar terrain for Lynch also, for example the ecological issues, the hippie ideals and the complex economic and political system of the planet. This created a mess rather the a tension. This doesn't mean that it doesn't conform to Lynch's vision, the deformity of the Baron, the dreams of Paul and the use of sound and speech are all very Lynchian touches but they only colour the film rather than structure it.
Lynch says (in 'Time' magazine), "I was making [Dune] for the producers not for myself. That's why the right of final cut is crucial. One person has to be a filter for everything. I believe this is a lesson world; we're supposed to learn stuff. But three and a half years to learn that lesson is too long." It is only through apocryphal stories that we can speculate on the amount of influence of producer and director. The producer of Dune was Raffaella de Laurentiis, daughter of Dino de Laurentiis, and she and Lynch reputedly didn't get on. In interview Lynch has talked of his vision of Dune as much longer, much more abstract and filmed in black and white. Many of the more 'gory' scenes had to be cut for cinematic release also.
Of The Elephant Man, Lynch says "I think I did the best job that could be done. And I didn't compromise except maybe two or three times, but no more than that" (Chion's 'David Lynch'). This really seems a great defence for the auteur theory. The Lynchian film was the success and the compromised vision was the failure. But, as Chion points out, "The Elephant man is also a great film made collectively", it would be myopic not to credit the performances of Anthony Hopkins and especially John Hurt. Poor special effects particularly mar Dune. Many more factors contributed to the fairing of both films than simply the direction. Lynch says (quoted in Chion's book), "The director is like a filter. Everything passes through me. Everybody has an input and has ideas and so the movie has a great momentum going for it. Some things pass right through the filter and some things don't".
In "Towards a Theory of Film History", Sarris states that "To look at a film as the expression of a director's vision is not to credit the director with total creativity. All directors, and not just in Hollywood, are imprisoned by the conditions of their craft and their culture". This would imply that a director's vision is never truly realised as a pure expression of his personality, because it cannot escape these cinematic and cultural conditions. In that case even our "control sample", Eraserhead, is tainted by an excess of outside influence. But it is obvious that the restrictions Lynch had to work within were radically different from what he encountered working on major Hollywood productions such as The Elephant Man and Dune. To continue this line of reasoning, if we could define these variable constraints into a form more theoretically tangible, to subtract them from any film within a director's oeuvre should, according to the auteur theory, leave us with a constant roughly representative of the director's personality, or the artistic expression of it. If this theoretical constant exists in all a director's films (as his personality is assumed to be invariable) it further implies that each auteur's film, according to the auteur theory, is of equal merit when divorced of its mode and environment of production. In that case for The Elephant Man to be an unmitigated success, while Dune was a resounding failure would imply the fault lay in the production. This tenuous line of reasoning conveniently exonerates Mr. Lynch of the failings of Dune but also denies him credit for the success of The Elephant Man. Of course, such a theoretic line is unrealistic as the director within the Hollywood system cannot be cleanly isolated from the production. The director is meshed within a production, the final product of a collaboration between a producer and director is ultimately destined to occupy the middle ground of compromise.
Robin Wood, in his 1989 introduction to 'Hitchcock's Films Revisited' makes a comment on Blue Velvet, which is interesting not for what it claims Blue Velvet achieved but for what it implies his two earlier films didn't;
" Like all Lynch's work it is highly idiosyncratic with all the stylistic signifiers of a 'personal statement'. Unlike The Elephant Man and Dune, it seems a work of someone - at least on a superficial level - completely in control of his material : the effects, that is to say, achieved through dialogue, acting, mise-en-scene, decor, composition, editing, etc., they have a precision that convinces one that they are exactly what Lynch wanted."
But Pam Cook, in 'The Cinema Book' puts it into context:
"Auteur theory enables the critic to impose unity in retrospect on a body of films produced in a variety of production set-ups. This unity is attributed to the presence of the director as the essential source of meaning, across different genres, studios etc. Ö The director as auteur is therefore opposed to Hollywood production conditions, and it is from this relationship of opposition that true authorship arises."
The Star System and Authorship
Robin Wood, in a later chapter of his book, goes on to say "Arguably the most important recent development in film theory has been the radical opening up of the discussion of stars: the construction of the persona, the intricate inter-relation of acting/presence/image, the way in which a star functions, and the complex meanings she or he generates within a given filmic text."
Making films within the Hollywood system automatically implies the influence of the star system. Richard Dyer in his book "Stars" pioneered the star theory, Wood proposes the star system as an opponent of auteur theory. Dyer examined how a star persona is constructed through promotion, publicity, their films and their commentaries and how this hybrid persona's presence in a work can undermine the generic or auteuristic readings of a film. Wood refers to this citing the example of Ingrid Bergman in Notorious(1946). Bergman's star persona had been built up as the wholesome, essentially good hearted lady, Wood defines it in four components; "nature and health; niceness; the lady; the actress". When Bergman appeared in Notorious she is initially presented against type, as drunken, cynical and promiscuous, but, Wood says "simply because she is irreducibly Bergman - we trust her". This implies that Hitchcock, intending a reversal of character in the course of the film is pre-empted by his audience due to Bergman's star persona. This further implies that the star persona has undermined Hitchcock's authorial intentionality.
This draws attention to the point that the star persona is an element of production that can influence a film's reading and therefore should be under control of the auteur if he is to retain his authorship. Lynch has consistently made use of actors to emphasise rather than undermine his authorship. He has a regular stable of favourites who appear repeatedly in his films, such as Kyle MacLachlan, Freddie Jones, Sherilyn Fenn, Frances Bay, Harry Dean Stanton, Sheryl Lee, and the late Jack Nance, who played the lead in Eraserhead and has appeared in every Lynch film since (although he ended up on the cutting room floor in Twin Peaks: Fire Walk With Me). Apart from his own stable, he also utilises "name" actors but uses elements of their star persona to enhance the role. For example, in Blue Velvet, the drug crazed psycho Frank was played by Denis Hopper, who bought to the role the echo of his Easy Rider (Hopper 1969) role 20 years earlier and also the history of his subsequent personal drug psychosis prior to Blue Velvet resurrecting his acting career. The camp pervert Ben has more resonance being played by ex-child-star Dean Stockwell. Isabella Rosselini, used as femme-fatale in both Blue Velvet and Wild At Heart, carries her unmistakable resemblance to her mother, the aforementioned Ingrid Bergman, and connotations that inspires. Rosellini is a personification of the subversion of her mother's star persona, she was the child of Bergman's affair with the director Roberto Rosellini that shocked fans familiar with the onscreen clean-cut persona. Michel Chion says "Her Italian looks are used as an exotic element, not by direct reference to an accent or an ethnicity, but by underlining how foreign a body she is in the neat little world of Lumberton".
This manipulation of the star system seems only to apply to his later films. In the case of Eraserhead, the budget would have been too low to allow recognisable faces to be used. With The Elephant Man and Dune, Many known actors are used to varying degrees of effectiveness. In The Elephant Man, Anthony Hopkins and Sir John Geilgud are cast very much to type and are very good in their roles. John Hurt carries the weight of his role entirely in the use of his voice, his face concealed by the extreme make-up the role required. Still, facing this limitation, he gives a stunning performance that is crucial to the pathos of the film. In Dune, stars such as Max von Sydow are reduced to simply cameos. A significant piece of casting in Dune, has to be the pop star Sting as the main villain. He was a very well known face at the time and appealed to a young audience but he is blatantly inappropriate in his role. It would be nice to attribute the casting of Sting as the responsibility of a producer targeting a specific market, but we know that Lynch had befriended Sting during his time at the ill-fated Zoetrope studios and so most likely influenced the casting of that role. In Sting's case though it wasn't his star persona that effected the role but rather his lack of one, having insufficient on screen presence to appear a formidable threat to the hero.
I cannot think of an example of a star's persona having a negative influence on the auteur's vision in the case of David Lynch, but above I mentioned some examples of a positive effect. It is probably too early into his career to formulate a theory on this matter, there are no examples of comparable works thus far (such as the Cary Grant/James Stewart films of Hitchcock), but Lynch, in his use of stars does seem aware of their influence on a text. The star system, instead of undermining the auteur is actually utilised as a tool by the auteur.
Structuralism and Genre
During the 1970's structuralism prompted debate over the auteur theory, the two theories seeming initially incompatible. As Pam Cook points out, "Structuralist criticism leads the critic away from conventional auteur criticism towards a consideration of other areas of cinema in an attempt to displace the auteur from the centre of the work. However, in the process the specificality of a particular director can be lost altogether". We can see how structuralism applies to cinema with Barthes's concepts of the "death of the author" and consequential "birth of the reader". Barthes states "the author does not write he is written", i.e. the author doesn't invent the language and conventions of his/her medium, the genre within which it is located or the ideological or historical location of production, the author is not free. This also applies to the reader of the text, their reading and interpretation is dependent on their reading environment, the presence of others, the gender, class or political leanings of the reader or even the reader's mood. Therefore it means an immutable meaning cannot be embedded in a text because the interpretation of a text can vary so radically. The ambiguity of meaning and consequent undermining of an author's intentionality is further discussed in Chapter 3.
The aforementioned constraints of language root in Saussure's theory of signification, the idea that "it's the language that speaks, not the author". We can see how this applies to the two films examined in this chapter, both of which adhere to genre convention. Genre serves to establish audience expectations and draws on an audiences accumulated knowledge, therefore the interpretation of genre comes from the reader. But, as Lapsley and Westlake say, "any system of rules brings with it the possibility of transgression, genre can be seen as providing a field for variation and elaboration of meaning". Rather than genre acting as an artistic straight-jacket the auteur can use genre as a tool. Just as it is the reader's duty to interpret genre conventions it the auteur's duty to subvert those preconceptions. Lynch incorporates a subversion of genre into his overall style, not only in his blatantly generic films (The Elephant Man, Dune, Wild At Heart) but also his less conventional narratives (Blue Velvet, Twin Peaks). Even in one of his earliest works, Six Figures(1967), Lynch refused to obey the conventions of the cinematic medium by using a three dimensional screen.
Lynch chooses a generic convention in his construction of a scene in much the same way as he would choose a lens to shoot it. Genre, in Lynch's work is expressive, not restrictive. John Orr, in 'Cinema and Modernity' says "Lynch is very contemporary precisely by returning to the past, the teller of parodic fables with endless echoes in his nation's film history. But the mocker of genre is also it's prisoner, worshipping with small town nostalgia the American dream which his films undercut." It is a thin line between the parodist and the prisoner, the structuralists state that it is the dog that wags the tail, while the auteurists insist it is the tail that wags the dog.
Lynch, quoted by Chion, has said "each film is different. Whatever genre it's in, as soon as you start a film, you're inside a setting and you have to stay in it". He is implying that cinema as a medium is a constraint regardless of the genre of a film. He acknowledges the structuralist proposal of the artist as a prisoner of his medium. It is important to remember that Lynch was a painter by origin, not a film-maker. The conventions of the two-dimensional canvas are even more restrictive than the moving narrative of cinema. In painting, the value of the art lies in its interpretation. Lynch's films are created with this philosophy, best demonstrated with the polysemic Eraserhead, if ever a film's interpretation lay with the reader it was there. But it is still an auteur film, it presents a world view but doesn't dictate how we should accept that view. Lynch's view is presented and it is up to the reader to incorporate that world view into their own. The structuralist critisism seems to focus on the constraints of the medium as undermining an author's 'message'. This implies a conscious embedding of meaning within a work, this is not necessarily a criterion of an auteur. A director's personality becomes apparent in a text not by construction (although this is possible as discussed in Chapter 4). This is similar to what Peter Wollen argues in his proposal for 'auteur-structuralism' in 'Signs and Meanings in Cinema'. Instead of uncovering an auteur within a text it is a structure within the text that is uncovered by auteur analysis, he says "Fuller or Hawks or Hitchcock, the directors are quite separate from 'Fuller' or 'Hawks' or 'Hitchcock', the structures named after them". Sarris acknowledges that the meanings within a text are as likely to be unconscious as they are conscious, Wollen takes this one step further with the film auteur becoming an "unconscious catalyst", the "bearer" of the structure rather than the creator, the binding element of commerce, script, cast etc. Lapsley/Westlake say:
"the method, Wollen cautioned, could not be content with simply identifying the diagnostic features that all the films of a given director had in common, it must also discover what distinguishes each film from the others, as well as repetition there must also be an awareness of difference; as well as universality, singularity."
Therefore, it is what makes them unique that is important in determining the auteur. The idea of the structure within the text fits in nicely with the idea of the constant that can be theoretically extracted from the film, as discussed earlier. Wollen talks of the need to remove the "noise" from producer, cameraman, actors etc. to be able to grasp the auteur structure. Wollen forms an uneasy compromise between the two theories, managing to simultaneously acknowledge the individual as the source of artistic production and also the prisoner of linguistic and social constructions.