• 1 See for example David Thomson’s comment in “Du Maurier, Hitchcock and Holding an Audience,” in Hele (...)
• 2 In the Internet Movie Database, Rebecca has been rated by more than 110,000 users and is by far the (...)

1Ever since she appeared as the second Mrs. de Winter in Alfred Hitchcock’s adaptation of Daphne Du Maurier’s Rebecca (1940), Joan Fontaine has been considered a perfect match for the role.1 The film immediately turned her into a top Hollywood star, and brought her critical acclaim and her first Academy Award nomination. It also typecast her as a clumsy, naïve girl transformed into a mature, composed woman, which she later successfully repeated in Suspicion (1941), The Constant Nymph (1943), Jane Eyre (1944) and A Letter from an Unknown Woman (1948). Today, some eighty years after the film’s release, it remains her most popular and iconic part.2 At the time of her casting, though, Joan Fontaine was an undistinguished and relatively unknown Hollywood performer and far from being the first or only choice for the demanding role. On the contrary, many American and British actresses – both experienced stars and novices – were considered and tested before the final decision was made.

• 3 John O. Thompson, “Screen Acting and the Commutation Test,” Screen, vol. 19, no. 2, July 1978, 55-7 (...)
• 4 All of these casting choices were in fact considered at one point or another. A whole book by Eila (...)

2It is always intriguing to speculate how a film’s impact and meaning would change if some other actor or actress had been cast in the leading role. More than just a harmless game, this intellectual exercise – inspired by the commutation test in structural linguistics in which one element in a signifying system is replaced by another3 – makes us more aware of the unique qualities each performer possesses and mobilizes in order to impersonate a character on the screen. Try to picture any canonical film with a different cast. For example, what would happen if Shirley Temple instead of Judy Garland played Dorothy in The Wizard of Oz (1939)? Would The Third Man (1949) be the same if Cary Grant were to play the male lead instead of Joseph Cotten? And, to name a more recent example, what if Emma Watson and Emma Stone were to switch their roles in Beauty and the Beast (2017) and La La Land (2016)?4

• 5 Thompson, “Screen Acting and the Commutation Test,” 67.

3Writing in 1978, John O. Thompson opined that “[t]here is room for a great deal of detailed research on the history of casting. […] It would be good to have accounts of actual casting practice detailed enough to serve as a control on the intuitions commutation affords us about possible and impossible matchings of actor to role.”5 In the case of Rebecca, we have at our disposal not only voluminous written documentation from producer David O. Selznick’s archive, but also actual screen tests that were made of the most serious contenders for the female lead. This means that we do not have to merely imagine alternatives to Joan Fontaine as the second Mrs. de Winter – the footage that has survived allows us to compare her creation with other aspirants while they were all “in character.”

• 6 David T. Johnson, “Adaptation and Fidelity,” in Thomas Leitch (ed.), The Oxford Handbook of Adaptat (...)
• 7 Guerric DeBona, Film Adaptation in the Studio Era, Urbana: University of Illinois Press, 2010, 1-11 (...)
• 8 Simone Murray, The Adaptation Industry: The Cultural Economy of Contemporary Literary Adaptation, N (...)

5Because I treat casting as an integral part of the adaptation process and repeatedly make use of the term “fidelity” – one of the most problematic notions in adaptation studies – a short explanatory note is necessary. Fidelity criticism (or fidelity studies) was once a mainstay of the discipline but in the past twenty or twenty-five years, we have witnessed a methodological shift which turned fidelity into a dated concept that is often met with disdain and abhorrence.6 This article on Selznick and Hitchcock’s version of Rebecca is certainly not meant as an intervention in the discussion on the possible merits or weaknesses of this once widespread approach. I am not even concerned with a detailed comparison of the film with the novel, nor do I wish to make evaluative claims about their respective qualities or shortcomings. On the contrary, inspired by Guerric DeBona’s pioneering work on film adaptations in the Hollywood studio era, I move away from a text-based approach to film adaptation to a more historically-grounded work which is preoccupied with the issues of agency, industrial choice and audience reception.7 In other words, formalism is substituted with a production- and industry-based approach investigating the specific material conditions (in this case the process of casting) which determined the final shape of the film artifact. As put forward by Simone Murray, adaptation is thus treated not “as an exercise in comparative textual analysis of individual print works and their screen versions, but as a material phenomenon produced by a system of interlinked interests and actors.”8 To sum up, in what follows, I approach fidelity not as a method but rather as an object of my study: as will be seen hereafter, for Selznick, “fidelity” was a deliberate goal which informed his adaptation, casting and marketing strategies.

• 9 Kyle Dawson Edwards discusses the brand Selznick created for his company in “Brand-Name Literature: (...)

6Both as a studio producer and as an independent, David O. Selznick was known to industry insiders and the general public as an adaptor of literary classics and, later in his career, of current bestsellers. His reputation as a master filmmaker with a consistently high level of artistic achievement rested on high-profile pictures derived from literary sources. His company’s motto, “in the tradition of quality,” determined the choice of material and style for his adaptations, which included Little Lord Fauntleroy, The Garden of Allah (both 1936), The Prisoner of Zenda (1937), The Adventures of Tom Sawyer (1938), Gone with the Wind (1939) and, earlier, while working at MGM, David Copperfield, Anna Karenina and A Tale of Two Cities (all 1935).9 A screen adaptation supervised by Selznick usually implied an esteemed, culturally-valued material, high production values, a stellar cast and a proclaimed fidelity to the source novel, motivated by the attempt to preserve all those qualities – however vaguely defined – that made the original a success. The goal was to create an overall aura of prestige and respectability that would in turn translate into good box office.

• 10 Quoted in Rudy Behlmer (ed.), Memo from David O. Selznick, New York: The Viking Press, 1972, 148, 1 (...)
• 11 Ibid., 149, 150.
• 12 Quoted in Steve Wilson, The Making of Gone with the Wind, Austin: University of Texas Press, 2014, (...)

7Identical standards were supposed to shape Selznick’s version of Daphne Du Maurier’s Rebecca, to which Alfred Hitchcock was assigned as director. The film was highly anticipated by audiences and critics alike as it was announced as the immediate successor of the Civil War epic Gone with the Wind which had been in the making since mid-1936. While adapting Margaret Mitchell’s bestselling novel, Selznick formulated several principles which guided the whole process. He was convinced that “it is much better to chop out whole sequences than it is to make small deletions in individual scenes or sequences […]. I don’t think there is much harm in rearranging sequences so long as the sequences are as the readers remember them and so long as cuts in these sequences are made so carefully that the losses are not discernible.”10 He also felt that “we will be forgiven for cuts if we do not invent sequences” and that “we should not attempt to correct seeming faults of construction. I have learned to avoid trying to improve on success.”11 Selznick was of the opinion that “one never knows what chemicals have gone to make up success and that a successful adaptation is obtained by the most conscientious effort to recapture these chemicals.”12

• 13 Ibid., 21. The making of Gone with the Wind and the challenging process of adapting Mitchell’s nov (...)
• 14 For an excellent account of the film’s production history, see Leonard Leff, Hitchcock & Selznick: (...)
• 15 Susan Ohmer, George Gallup in Hollywood, New York: Columbia University Press, 2006, 128.
• 16 Rudy Behlmer, op. cit., 269.

8Basically, Selznick urged his screenwriter Sidney Howard, who was entrusted with writing the first draft of the screenplay for Gone with the Wind, as well as his numerous successors, to stay as close to the novel as possible. The writers were authorized to make necessary eliminations and reductions (these were unavoidable in most film adaptations and particularly in this unprecedented attempt to transform a 1,000-plus-page novel into a film of reasonable length) but Selznick discouraged them from inventing new scenes or making undesirable alterations to something which was proven so successful in the past. As he emphasized in a letter early in the screenwriting process, “I do not share the general motion picture theory that the screen is a different medium and because of that necessitates a different approach. I feel that the screen is much closer to the novel in form than it is to the play and that its latitude is at least that of the novel.”13 Similar concerns, amplified – as was the case with Gone with the Wind – by the recentness of the novel and its popular success, emerged when Hitchcock and his collaborators Joan Harrison and Philip MacDonald (with the assistance of the director’s wife Alma Reville) started working on the screenplay for Rebecca.14 After delivering a detailed, ninety-page treatment in June 1939, Selznick, furious, responded with the oft-quoted words “We bought REBECCA, and we intend to make REBECCA.” He called it “a distorted and vulgarized version of a provenly successful work” and went on to elaborate on the reasons for his dismissal of the treatment. Specifically, the producer rejected a drastic change of exposition in Monte Carlo; sharply dismissed the idea to give “I” a name (“Daphne”); and insisted upon building a sense of empathy and identification between her and members of the audience which, according to the Audience Research Institute, was projected to be 71% female.15 He also didn’t care for Hitchcock’s attempts to insert humor into the story and, on the contrary, in line with the previous point, insisted on preserving all the little details that deepened characterization or illustrated “I’s” thoughts and emotions: “We have removed all the subtleties and substituted big broad strokes which in outline form betray just how ordinary the actual plot is and just how bad a picture it would make without the little feminine things which are so recognizable and which make every woman say, ‘I know just how she feels… I know just what she’s going through…’ etc.”16

• 17 Ibid., 266.

9As in his earlier adaptations, Selznick demanded fidelity and rejected any substitutions of the screenwriters’ invention: “Readers of a dearly loved book will forgive omissions if there is an obvious reason for them; but very properly, they will not forgive substitutions.” The producer believed that the only grounds for making justifiable omissions were “length, censorship, or other practical considerations,” presumably technical difficulties or unreasonable expenses.17 In a detailed memo to Hitchcock dated June 12, 1939, Selznick then summed up his theory of adaptation, which – he wasn’t shy to emphasize – was based on a proven track record of past successes:

• 19 Kyle Dawson Edwards, “Brand-Name Literature: Film Adaptation and Selznick International Pictures’ R (...)

10In other words, as pointed out by Kyle Dawson Edwards, Selznick argued that “film adaptations should reproduce and preserve rather than interpret their sources.”19

• 20 For the same reason, the title of the book was preserved even though there were some objections aga (...)

11Selznick used his authority as the film’s producer and Hitchcock’s employer to order a completely new treatment that would correct these shortcomings and be more faithful to the novel. It is perhaps useful to reiterate that the producer’s motivation for taking such an approach to adapting Rebecca was both of an artistic and a commercial nature: according to Selznick, popular material demanded fidelity (it wouldn’t so much matter in adaptations of unsuccessful works) which reduced the risk inherent in film production and made for good box office. For instance, not giving the main protagonist a name was for Selznick not only a matter of storytelling but also of showmanship: this is how the readers remembered the book and they would no doubt expect this element preserved in the film version.20

• 21 Rudy Behlmer, op. cit., 279.

12I have included this lengthy prelude about Selznick’s approach to film adaptation because I am convinced that the same assumptions and concerns guided the casting of Rebecca. In the producer’s view, only an attractive cast true to Du Maurier’s characterizations would translate into a popular (and thus commercial) success. However, despite Selznick’s call for “accurate casting,”21 the process of selecting suitable performers for the lead roles is not an exact science. Quite on the contrary, a lot of uncertainty is typically involved.

• 22 The process of casting the role of the second Mrs. de Winter in Rebecca is covered in other books b (...)
• 23 The talent hunt is covered by Rupert Alistair in The Search for Scarlett O’Hara: Gone with the Wind (...)
• 24 In addition to Gone with the Wind and Rebecca, Selznick mounted big, nation-wide talent searches wh (...)
• 25 Thomas Schatz, Boom and Bust: The American Cinema in the 1940s, New York: Charles Scribner’s Sons, (...)

13As suggested, it was the requirement for “fidelity” to Du Maurier’s literary original that also informed the careful process of casting the main roles. For Selznick, this had always been an important consideration in a film’s development. For example, it took him more than two years to fill the role of Scarlett O’Hara in Gone with the Wind and even though the talent search was extremely costly and, as the shooting was about to start, quite nerve-wracking, it also made for great publicity.23 The process of casting Rebecca wasn’t as long and extensive (after all, Gone with the Wind was the greatest production of all time), but it still was very thorough and time-consuming when compared to industry standards.24 In the highly competitive American film industry, 300 to 400 films a year competed at the box office.25 Selznick, who usually produced only one or two prestige pictures a year, was well aware that an attractive cast was one of the best insurances against commercial failure and that it mattered particularly in adaptations of classic works and bestsellers because filmgoers’ expectations were high, and they tended to be dismissive when their conception of favorite characters didn’t match actual casting choices. Moreover, unlike MGM, Twentieth Century-Fox or Paramount, his company, Selznick International, had by the late 1930s only a limited talent roster and was therefore compelled to consider actors and actresses owned by other studios (while the majors would probably simply settle for one of their contract players). This meant that, in theory at least, more options were open for consideration, but those also often led to extended negotiations.

• 26 Rudy Behlmer, op. cit., 272.

14Finding the right actor to play Maxim de Winter, master of Manderley, was quite straightforward: after the refusal of Ronald Colman, called by Selznick “THE ONLY PERFECT MAN” for the part,26 the assignment went to Laurence Olivier, fresh from the success of Samuel Goldwyn and William Wyler’s production of Wuthering Heights (1939). It was the female lead that proved much more difficult to cast.

15The role of the second Mrs. de Winter was in great demand for a number of reasons: it was a complex character who developed from a mousy, inexperienced girl into a confident woman; the film was based on an international bestseller and being developed by one of Hollywood’s most respected filmmakers as his first production after Gone with the Wind (which was scheduled to be released in December 1939); and the director was to be the renowned Alfred Hitchcock, who had recently been imported to Hollywood from England after the signing of a seven-year contract with Selznick in March 1939. There was little doubt that, if cast right, the film could provide a tremendous boost to almost anyone’s career.

• 27 See Leonard Leff, , op. cit., 82.

17Between July 1938, when Selznick started negotiating for the adaptation rights, and September 1939, when the picture finally went before cameras, dozens of candidates were considered. They included established personalities and unknowns, both British (who would provide the “correct” accent and an aura of authenticity) and American (whose foreignness in terms of the fictional world was not necessarily perceived as an obstacle because it strengthened the character’s sense of being out of place at Manderley). The whole process was supervised by Selznick and Hitchcock with important input from others including Hitchcock’s secretary Joan Harrison, his wife Alma Reville, Selznick’s Eastern representative and talent scout Kay Brown, and John Hay Whitney, the primary investor in Selznick International. Selznick and Hitchcock had serious differences of opinion – just as with the screenplay and other aspects of the production – but in the end these seem to have had a positive effect: the two men challenged each other, which eventually led to better results.27

• 28 David O. Selznick, Telegram from 7 October 1938, Box 170, Folder 12, Rebecca, Casting 2nd Wife, Dav (...)
• 29 David O. Selznick, Memo to Alfred Hitchcock from 29 June 1939, Box 951, Folder 4, Rebecca Casting a (...)

18London-born actress Nova Pilbeam was the first serious candidate for “I.” She had starred in two Alfred Hitchcock films – The Man Who Knew Too Much (1934) and Young and Innocent (1937) – but, surprisingly, it was Selznick who became her champion, considering even signing a long-term contract with her and turning her into a big star under his management. Hitchcock, however, was fundamentally opposed to her because he was convinced that she lacked humor, had a narrow range and was “TOO IMMATURE AND DIFFICULT [TO] HANDLE IN LOVE SCENES.”28 Instead, he suggested Margaret Lockwood, whom he directed in The Lady Vanishes (1938). Pilbeam’s and Lockwood’s names would be mentioned as remote possibilities in the months to come but they were soon joined by a legion of other hopefuls. Until spring 1939 the exchange of suggestions between Selznick, Hitchcock and others was lively and uninhibited but beginning in April and May, the process got more rigorous. Hopeless cases, inappropriate for the role, were rejected outright based on their photographs only. Others were invited for readings and only the most promising candidates were called in again for screen tests, which were regarded as the most sophisticated and reliable means of making the final decision. Screen tests of the highest technical standard were quite costly, but Selznick wanted to explore all possibilies and was prepared to spare a generous amount of money on them. He told Hitchcock that “I don’t care if you test twenty girls, provided only that we don’t waste money on obviously poor possibilities; that we weed out what we can through readings; and that we organize the tests to do as many possible in a day and thereby hold down the cost.”29

• 31 De Havilland had a long-term contract at Warner Bros., but Goldwyn obtained an option for one film (...)

20After several weeks, the selection was narrowed down to six contenders – Anne Baxter, Olivia de Havilland, Joan Fontaine, Vivien Leigh, Margaret Sullavan and Loretta Young. Of these, de Havilland was shortly ruled out – partly because getting her would mean dealing with the notoriously difficult studio moguls Jack Warner and Samuel Goldwyn31, and partly because it would hurt the negotiations with her sister Joan Fontaine with whom she had an uneasy relationship. The remaining five actresses underwent further tests to determine who might fill the role of the second Mrs. de Winter most competently.

• 32 Jeanine Basinger, The Star Machine, New York: Alfred A. Knopf, 2007, 39.

21A short note on screen tests – according to Jeanine Basinger “a largely unexplored area of movie history”32 – is in order. In her book The Star Machine, Basinger distinguishes between three types of tests:

• 34 As Jeanine Basinger reminds us, “people don’t always look on film the way they look in real life” ((...)
• 35 The scene as used in the screen tests is different from the finished film: the dialogue is longer a (...)
• 36 James Naremore, Acting in the Cinema, Berkeley, Los Angeles and London: University of California Pr (...)

23The screen tests that Hitchcock and Selznick supervised were not intended for public consumption; their purpose was to assist them in evaluating the appropriateness of casting by matching the performance of an actress with their conception of the character and by comparing it with other candidates. In general, three sets of data were analyzed: the performer’s appearance (her physiognomy and how she photographed on film34); her personality (how she responded to the camera and which qualities of her “came through”); and her performance choices (the gestures and actions she used to interpret the scenario). In order to make the best decision possible, screen tests had to be of high quality and had to provide each candidate with comparable conditions in terms of scene selection, cameraman, lighting, costume, make-up, screen partner, etc. Selznick and Hitchcock picked the scene where the newly wedded couple, Maxim and “I”, clash over the broken china cupid which had once belonged to Rebecca. The choice seems well-founded since the scene is very dramatic, heavy on dialogue and reaction, and provides ample opportunity for displaying nervousness and uneasiness but also a hint of assertiveness, all qualities which were deemed essential for an effective portrayal of the lead character.35 The scene was played out without the benefit of continuity so the actresses had to concentrate their whole notion of the character into a few moments of drama. With the choice of the scene settled, Selznick and Hitchcock also made sure that the five aspirants for the role had flattering costumes (customized to each of them) and suitable acting partners (Alan Marshal, Reginald Denny, John Burton and Laurence Olivier, all skillful professionals, were used in the tests). Most of the tests were conducted in July and August 1939, with the final decision postponed until three weeks before the shooting was scheduled to start. Vivien Leigh, Selznick’s Scarlett in Gone with the Wind, was enthusiastic about the prospect of appearing in Rebecca but was tested primarily as a courtesy to her lover Laurence Olivier who had already clinched the role of Maxim de Winter. Since early 1939, she was Selznick’s contract player, which meant that using her would be simple and inexpensive (she was receiving a weekly salary regardless of whether she worked or not). However, the consensus at the studio was that she was not the correct type for the fragile “I.” In her first test against Alan Marshal, Leigh seems flirtatious and bitchy rather than nervous and timid. When accepting the blame for the broken ornament, she crumples her sweater, supposedly to suggest uneasiness (thus turning it – in James Naremore’s words – into an expressive object36) but it looks as though she’s about to unbutton and undress. Later, when she says that she’s “dull and quiet and inexperienced,” it sharply contrasts with her appearance and conduct. To indicate her character’s defensiveness, Leigh uses a high-pitched voice and fast speech but the effect, once again, is of manipulative trickery rather than uneasiness. In her second test, Leigh was allowed to perform with Laurence Olivier and the result was an improvement. This time, she seems more in character and her “I” does come across as a shy and insecure creature. Leigh’s performance benefited from being able to move around the set a little which allowed her to express emotions with posture and small gestures (in her first test, she was seated behind a table). Nevertheless, it still wasn’t enough to convince Selznick and Hitchcock that she was suitable for the part. As Selznick informed John Whitney, Leigh’s only defender at the studio,

• 38 Earlier the same year, Young was considered for a role in Selznick’s Intermezzo: A Love Story, but (...)
• 39 David Shipman, The Great Movie Stars 1: The Golden Years, London: Macdonald, 1989, 611.
• 40 David O. Selznick, Memo to Ray Klune from 6 October 1939, Box 170, Folder 9, Rebecca Cast I – Fonta (...)
• 41 Selznick, Memo to John Hay Whitney, 18 August 1939.

25Loretta Young, born in 1913 just like Vivien Leigh, was until recently a contract player at Twentieth Century-Fox where she specialized in romance and comedy, but in 1939 she started free-lancing. This proved more challenging than expected and a lead role in Rebecca seemed a welcome opportunity to get a solid footing in her new position in the industry.38 With her big eyes and apple cheeks, she was known as “Hollywood’s beautiful hack.”39 However, her radiant looks ultimately worked to her disadvantage. Selznick visualized “I” as an “unglamorous creature,” though “sufficiently pretty and appealing, in a simple girlish way, to understand why Maxim would marry her.”40 In the test, an attempt was made to deglamorize Young by dressing her in a dull sweater, but she still looks highly glamorous in close-ups. Regardless of her appearance, her performance in the test was not wholly convincing either: her reactions were slow and she kept on biting her lip to illustrate nervousness. In the final analysis she doesn’t seem shy, insecure or frightened but merely sad and disappointed as though her present disposition is just a passing mood, not a defining character trait. Moreover, Young’s strong American accent was not appropriate and overall, she was too much of an all-American girl to effectively portray a British companion. As aptly summed up by Selznick: “We have ruled out Loretta Young on the ground that when you strip her of Hollywood glamour, stunning clothes, a great deal of make-up, etc., there is very little left. […] Incidentally, her accent, of course, does her no good for the role. We are prepared for an American accent, if necessary, but not one quite so Santa Monica Boulevard as Loretta’s.”41

• 42 At one point, she was also seriously considered for the role of Scarlett O’Hara in Gone with the Wi (...)
• 43 Donald Spoto, Spellbound by Beauty: Alfred Hitchcock and His Leading Ladies, New York: Harmony Book (...)

26In the final trio, Margaret Sullavan, recently nominated for an Oscar for her impressive performance in Three Comrades (1938), was the oldest: she was 28 when the tests for Rebecca were made.42 In the footage, she fails to project innocence, fragility or naiveté. On the contrary, she seems “quirkily autonomous”,43 too sure of herself, as experienced and mature as Maxim (her costume – a white buttoned blouse with a high collar – contributes to this impression). Sullavan’s characteristically husky voice and rapid cadence of speech also contradict any attempts at invoking vulnerability. As opposed to Loretta Young, her reactions in the test are too fast: sometimes she even speaks before Reginald Denny as Maxim finishes his lines. In consequence, her confident conduct makes it hardly credible that her “I” wouldn’t be able to stand up to Maxim or Mrs. Danvers – for example, it is impossible to visualize Mrs. Danvers bringing her to the brink of suicide. Still, Sullavan was an accomplished actress and her performance never becomes monotonous; by varying her expression and speech, she manages to achieve some level of emotional intensity. Even so, her casting in the role would necessitate many changes in the screenplay (the least of which would be references to the girl’s age) and the whole concept of an uneven romance between an inexperienced ingenue and an older man would be ruined. There were also practical considerations that, in the end, prevented Sullavan from obtaining the role. Even though she was married to the influential talent agent Leland Hayward, who pushed hard to get her the assignment, she was at the same time contracted to MGM, whose representatives demanded a substantial sum of money for her services and a one-film option for Vivien Leigh, which Selznick opposed.

• 44 Kay Brown Telegram to David O. Selznick from 14 August 1939, Box 174, Folder 1, Rebecca Tests, Davi (...)
• 45 Selznick, Memo to John Hay Whitney, 18 August 1939.
• 46 Ibid.
• 47 Not long afterwards, Baxter signed a contract with Twentieth Century-Fox and debuted (on loanout to (...)

27In contrast to Sullavan, the sixteen-year-old Anne Baxter was the youngest of the lot. Even though she was new to the film industry, she had performed successfully on Broadway and in her screen test for Rebecca, she made a very strong impression. Kay Brown opined that she was “MOST TOUCHING AND BEAUTIFUL”44 and Selznick concurred, saying that she “had more sincerity than Fontaine.”45 In her interpretation of the scene, she projects many of the qualities that were considered essential for the part: innocence, vulnerability, insecurity. She emphasizes “I”’s submissiveness by a careful selection of inventive gestures: she lowers her eyes, thus evading Maxim’s look; nervously squeezes her knitting; and kneels beside Maxim to literally look up to him. As the scene progresses, she very touchingly seeks Maxim’s confirmation of being happy in their marriage and still later, she holds a fragment of the china cupid to suggest a failure to do so. All of these performance choices prove that despite her age, Baxter already was a complete, highly skilled actress on par with her more experienced colleagues. However, her young age still posed a big problem because her pairing with Olivier – who at 32 was twice her age – would not look acceptable. Selznick explained to John Whitney that Baxter “is ten times more difficult to photograph than Fontaine, and I think it is a little harder to understand Max de Winter marrying her than it would be Fontaine.”46 The filmmakers experimented with her looks but neither a new hairdress, nor a blonde wig could conquer the prevailing sentiment that she looked like a schoolgirl in adult woman’s clothes.47

• 48 B. R. C, “The Screen,” The New York Times, 25 June 1937, 25.
• 49 Quoted in Marsha Lynn Beeman, Joan Fontaine: A Bio-Bibliography, Westport, Connecticut: Greenwood P (...)
• 50 Ibid., 73.
• 51 Her most distinctive part to date was a supporting role in the MGM production of The Women (1939), (...)

28Leigh, Young, Sullavan and Baxter all had significant assets but, in each case, there was also something that prevented them from getting the approval of both Selznick and Hitchcock: Leigh seemed too sly and coquettish, Young couldn’t possibly pass for a British companion, Sullavan was too old and Baxter too immature. Only Joan Fontaine seemed to be the ideal type: at 21, she was the right age; she was photogenic and good-looking but not exactly breathtaking; and as a child of British parents, she had the “correct accent.” Unlike Leigh, Sullavan or even Baxter, though, her acting skills were often disputed. She debuted in 1935 in the Joan Crawford-Robert Montgomery star vehicle No More Ladies in which she was credited as Joan Burfield. After signing a long-term contract with RKO, she was subjected to a typical studio build-up (from walk-in parts to supporting roles to leads in B movies) which, however, didn’t produce much result. For example, of the low-budget comedy/romance You Can’t Beat Love (1937), The New York Times said that it “should be run through the reviewing mill as perfunctorily as possible.”48 The critic didn’t find Fontaine’s performance worthy of a comment, but the trade journal Variety noticed that she “has looks” but also “lots to learn about camera technique and vocal modulation.”49 Even when the quality of her films improved, Fontaine’s notices remained unimpressive. Variety said that in A Damsel in Distress (1937), a musical comedy starring Fred Astaire, “Joan Fontaine is passively fair as the ingenue, nicely looking the role but otherwise undistinguished.”50 With her career stalled, RKO released her from the contract and, in early 1939, she became a free-lancer.51

• 52 Quoted in David Thomson, Showman: The Life of David O. Selznick, New York: Alfred A. Knopf, 1992, 3 (...)
• 53 Alfred Hitchcock, Memo to David O. Selznick from 19 August 1939, Box 174, Folder 1, Rebecca Tests, (...)
• 54 Polly [?] to Mr. O, undated letter.

29Her tests for Rebecca were promising, though not entirely convincing. In the first series, Fontaine conveys her character’s insecurity and submissiveness by posture – she is unnaturally slouched and looks like a dog waiting to be punished. She says her lines as if in a trance, her eyes wandering and avoiding Maxim’s gaze. Later, she emphasizes “I”’s emotional strain by raising her hands and massaging her temples. In the second surviving test, Fontaine goes even further. She sobs, kneads her hands, roams her eyes from one place to another, breathes loudly – all of which transforms her agitation into impending hysteria. For some, this was unbearable overacting. John Whitney wrote to Selznick that “the last test of Joan Fontaine was so bad that I cannot see her playing the role otherwise than as a dithering idiot, or as her other version – a talking magazine cover.”52 Alfred Hitchcock, his wife Alma and Joan Harrison all thought that Fontaine “was just too coy and simpering to a degree that it was intolerable” and “her voice was irritating.”53 There was resistance against Fontaine from the public as well. In a letter addressed to “Mr. O” (Selznick or, possibly, Laurence Olivier), a woman named Polly writes that giving the role to Fontaine would “ruin [the film] utterly. […] She knows nothing about acting… Any waitress without an ounce of experience could do it – even I!”54

• 55 David O. Selznick, Memo to Daniel O’Shea from 26 April 1938, Box 3338, Folder 9, Joan Fontaine Tale (...)
• 56 Later, Hitchcock would tell Franҫois Truffaut that he preferred Fontaine all along (François Truffa (...)
• 57 David O. Selznick, Telegram from 5 September 1939, Box 170, Folder 11, Rebecca, Casting 2nd Wife, D (...)

30Despite strong opposition, Selznick became Fontaine’s champion, privileging her appropriate type over any shortcomings she might have had as an actress. In other words, he preferred that the character of “I” – innocent, shy, awkward, gauche, unassuming – should be created not as an effect of acting or impersonation, but rather as a result of selecting a girl with proper physical features and on-screen presence. Already in spring 1938, he noticed that Fontaine “has a completely fresh quality that might be built with the right parts to stardom. She’s going to get nowhere at RKO.”55 Sixteen months later, Selznick, still believing in her potential, used his authority as the producer of Rebecca to prevail on Hitchcock (who, despite some reservations, preferred Sullavan56), Whitney (who supported Leigh) and other sceptics at the studio, signing a seven-year contract with Fontaine, with Rebecca as her first assignment. Eschewing the slow, but ultimately meaningless build-up she underwent at RKO, under Selznick’s supervision and guidance she was immediately cast in a coveted and difficult role in a prestigious, high-budget adaptation of a literary bestseller. The studio prepared a press release, informing the media that “THIS MAKES HER THE THIRD COMPARATIVE UNKNOWN ON WHOM [DAVID O. SELZNICK] HAS BEEN WILLING TO GAMBLE A STAR ASSIGNMENT DURING PAST YEAR – VIVIEN LEIGH AS SCARLETT, INGRID BERGMAN, SWEDISH STAR IN ‘INTERMEZZO: A LOVE STORY’ AND FONTAINE.”57

• 58 Selznick to John Hay Whitney, 18 August 1939.
• 59 According to Donald Spoto, sometimes he achieved this using questionable methods (Spellbound by Bea (...)
• 60 Leonard Leff, op. cit., 80.

31Even though Fontaine convinced Selznick that she knew “what the part is all about”,58 it was obvious that she would need careful direction. Hitchcock spent an uncharacteristically large amount of time coaching her to vary her performance, so it wouldn’t become monotonous, and guiding her through the transformation from a mousy girl to a more self-confident wife and mistress of the house.59 Despite the extraordinary care, numerous imperfections had to be solved with retakes and in postproduction. As remarked by Leonard Leff, “while the picture was not made in the editing and dubbing rooms, the increasingly favorable response of preview audiences, especially to Joan Fontaine’s performance, indicates that it was remarkably enhanced there.”60 This was all result of privileging an actress who was less accomplished and experienced but who provided her character with appropriate qualities in terms of appearance and personality – just as Selznick’s “theory of adaptation” required.

• 61 Dan Auiler, op. cit., 479. 7% of the respondents provided no answer and three of them said that Fon (...)

32The risky bet on Joan Fontaine – up to then regarded as an actress without much star-potential – paid off. The film was an unambiguous success. Fontaine was heralded as a fresh find (despite having made fifteen films in the past five years) and Selznick also won his share of plaudits as the discoverer who provided her with her first real breakthrough. The reactions of preview audiences already suggested that Fontaine’s interpretation of “I” met with approval. In a poll conducted after the preview screening of 25 December 1939, 53% of the respondents indicated that they “will go to see other pictures because Joan Fontaine is in them.” Another 35% stated that they liked her performance but will not go to her next film just because she’s in it. Ten respondents (or 4%) out of the total of 269 said that they “don’t care one way or the other” and no one was opposed to the actress to the extent that they wouldn’t go to see a picture with her.61

• 64 Rebecca advertisement in Motion Picture Herald, 16 March 1940.

35With high-brow literary adaptations, the approval of critics was always welcome because it lent the product much-needed prestige, but what ultimately mattered most was the reaction of paying audiences. After all, there were thousands of those who had loved the novel and whose conception of the story as well as the main character might have been markedly different from what the film – by its nature a very explicit medium – offered. The success of Selznick’s strategy, then, depended to a large extent on how effectively the idea of fidelity would be sold to the public. One only needs to have a look at the theatrical trailer to see that the film’s promotion was heavily based on the assurance that it was indeed a faithful translation of the novel. Scenes from the picture are intercut with pages from the book with highlighted lines spoken verbatim by Joan Fontaine’s “I” in the audio track. The narrator’s authoritative voice dispels all doubt by promising that the novel was “brought to the screen with all the warmth and emotion that made millions of readers acclaim Daphne du Maurier’s bestseller as the most exciting love story of our time.” Similarly, posters and numerous newspaper and magazine advertisements included a reproduction of the book’s cover (next to the likenesses of Joan Fontaine and Laurence Olivier) and promised another cinematic triumph “produced with equal faithfulness by the organization that gave you ‘Gone with the Wind’.”64 Moreover, prior to the film’s premiere, the well-publicized talent search made also distinctly clear that considerable personal and financial resources were invested in order to select the most suitable candidate for “I.”

• 65 Walt, “Rebecca,” 17; Thomas Schatz, Boom and Bust: The American Cinema in the 1940s, New York: Char (...)
• 66 “Rebecca Wins Critics’ Poll” 1.
• 67 Lupton Wilkinson, “Next Year’s Academy Award Winner!,” Hollywood, vol. 30, no. 7, July 1941, 19.

36On its first release, Rebecca grossed $3,000,000, finishing among the top 5 box office champions for 1940 despite predictions that it had a limited audience appeal.65 The film was chosen by American critics as the best picture of the year, beating The Grapes of Wrath, Ninotchka and Hitchcock’s own Foreign Correspondent,66 and went on to win two Academy Awards for Best Picture and Best Cinematography. Joan Fontaine was nominated in the Best Actress category but lost to Ginger Rogers for Kitty Foyle: The Natural History of a Woman (1940). When she won the following year for her performance in Hitchcock’s Suspicion (1941), many commentators credited Rebecca for securing her the award.67