In our paper “Defining film noir”, David Cornwell and I suggested it would be fruitful to systematically compare films we had identified as being typically described as “film noir” with films not appearing in any lists of film noir. Our first attempt to follow this through was the main content of two conference presentations, “Defining film noir 2”, for the Popular Culture Association in 1995, and “American film noir: Techniques and topics” for the Scottish Word and Image Group in 1998, but not subsequently in print. In repeating the results here, I am not suggesting that this study resulted in any definitive findings. However, I do think it illustrates the method we were proposing. As well as describing what we did and what we found, I shall note certain weaknesses in this initial stab at an empirical study.
A first issue to decide was what in practice would be a useful criterion of a film noir for such a study. We decided that films scoring between 10 and 12 in our master list would be the most likely to demonstrate the key features of film noir. We further decided to make use only of films to which we had access for repeated re-examination. This was the 1990s so we were employing VHS recordings. Not owning copies of all films meeting this criterion, we created a sample drawing on the films available to us which roughly corresponded in time to the distribution of 10-12 scoring films (called here CH10, CH11, CH12). We then looked for non-noir films which were released at approximately the same times. Since our focus period was 1941 to 1958, a convenient way of dividing it was in blocks of three years. Table 1 shows how the CH10-12 films were distributed over time and the two samples to be studied. It might be questioned how representative the non-noir sample is of non-noir films of the time generally. The issue of availability is obviously relevant here and those we made use of were available to us largely because we had collected them for other purposes. If our general procedure is indeed considered fruitful, the study could be repeated with a sample created by other means.
TABLE 1 TIME SPREAD OF FILMS
|DATES||ALL FILM NOIR||FILM NOIR SAMPLE||NON-NOIR SAMPLE|
The five characteristics which we explored were of differing sorts and could give rise to a variety of issues.
First we looked at whether the film derived from a novel. The influence of authors such as Hammett, Chandler and Cain is often remarked upon in discussions of film noir. Ginette Vincendeau when delineating the films she is discussing includes the phrase “often based on crime literature” (1992, p 50). We did not have a definition of “crime literature” but it seemed to us appropriate as a first step to consider novels generally as a source. If they could be shown to be common, then a further check could be made on the type of novel concerned.
Secondly, is the film in colour or black and white?
Thirdly, we looked at whether the film relies on voice-over narration. This was a check on our casual observation that several noir films did employ this technique. Sarah Kozloff, in her book Invisible Storytellers (1989, p. 34), argues that the 1940s was the “heyday” of voice-over in Hollywood films and cites film noir as a genre which exemplifies this. She also argues (p. 36) that in the 1950s the flood of voice-over narration “subsided”. Since our films cover both the 1940s and the 1950s, we were in a position to throw light on this claim.
Fourth, we considered whether the film is predominantly set in contemporary urban America. As we shall see, this was a particularly clear aspect of the difference between noir and non-noir films. However, I shall also suggest that a more nuanced exploration of this area might have been appropriate.
Fifth, to follow up Vernet’s question of “why the genre of films often called noir begins with an air of quietude” (1983, p 3), we examined whether the films had tense openings.
TABLE 2 SUMMARY OF COMPARISONS
|FACTORS||42 NOIR||42 NON-NOIR|
|Based on novel||21||15|
|Contemporary Urban America||32||14|
Obviously these characteristics need to be studied in different ways. Origin in a novel does not in fact require the actual film to be available (although it is helpful if the credits give the origin). Other characteristics considered not only required us to watch the films, but also gave rise to judgments which might be questioned. What counts as ”tense”, for example? However, we felt justified in exploring these interesting questions, because we would be using our judgments in no more subjective a way than is generally found in the literature of this field.
The results of our study are summarized in Table 3. Contrary to Marc Vernet’s claim, we found most of the CH10-12 films had a tense opening rather than “quietude”, tense openings being quite common too in the non-noir sample. Exactly half of the film noir sample were based on a novel, slightly more than was the case with the comparison group. None of our noir sample were in colour, which is not surprising since some list compilers appear to regard colour as necessarily excluding a film from being called “noir”, though it should be noted that few of the non-noir group were in colour either. Early voice-over narrative was not particularly common, and less frequent in the noir than in the non-noir group. The biggest contrast between the two samples was with respect to the setting. A substantial majority of the CH10-12 group had a contemporary urban American setting, compared to only a third of the non-noir comparison group. Some of these results were as we predicted but others were not, demonstrating that casual impressions do not necessarily stand up to more systematic investigation.
TABLE 3 COMPARISONS
1. Based on Novel
2. Colour/Partial colour
3. Early Narration: Voice-over
4. Contemporary, Urban, America predominant setting
5. Tense opening
|N||–||–||A||–||The Maltese Falcon||–||–||–||–||–||The Major and the Minor|
|N||–||–||A||–||The Big Sleep||N||–||–||–||–||Address Unknown|
|–||–||–||A||T||The Blue Dahlia||–||–||–||–||T||Bermuda Mystery|
|–||–||–||–||–||Conflict||N||–||V||A||T||Betrayal from the East|
|–||–||–||A||–||Dark Corner||–||–||–||–||–||The Body Snatcher|
|N||–||–||A||T||Fallen Angel||N||C||V||–||–||Duel in the Sun|
|N||–||V||A||T||Double Indemnity||–||–||V||–||–||The Enchanted Cottage|
|–||–||V||–||T||Gilda||–||–||–||A||–||The Falcon in Hollywood|
|N||–||V||A||T||Laura||N||–||V||A||–||From This Day Forward|
|N||–||–||A||T||Mildred Pierce||–||–||–||–||T||Isle of the Dead|
|N||–||V||A||T||“Murder, My Sweet”||N||–||–||A||–||It’s in the Bag|
|N||–||–||A||–||Phantom Lady||–||–||–||–||T||The Master Race|
|N||–||–||–||–||Scarlet Street||–||–||–||A||T||Murder in the Music Hall|
|–||–||–||–||T||The Strange Love of Martha Ivers||–||C||–||A||T||Wonder Man|
|N||–||–||A||–||The Woman in the Window||–||–||V||A||–||Youth Runs Wild|
|–||–||–||A||T||Body and Soul||–||–||–||–||–||The Beast With Five Fingers|
|–||–||V||A||T||Call Northside 777||N||–||–||–||T||The Ghost and Mrs Muir|
|N||–||–||A||T||Dark Passage||N||–||–||–||–||The Great Sinner|
|–||–||–||A||T||File on Thelma Jordon||–||–||–||–||T||Intrigue|
|–||–||–||A||–||Johnny O’Clock||N||–||V||–||T||Letter from a Unknown Woman|
|–||–||–||–||T||Key Largo||N||–||V||A||–||Letter to Three Wives|
|–||–||V||A||T||Kiss of Death||–||–||–||–||T||Macbeth|
|N||–||–||A||–||Lady in the Lake||–||–||–||A||–||Magic Town|
|N||–||–||–||–||Nightmare Alley||–||C||–||–||–||The Pirate|
|N||–||–||–||–||Out of the Past||N||P||–||–||–||The Private Affairs of Bel Ami|
|–||–||–||A||–||Set-Up||–||–||V||–||T||Sands of Iwo Jima|
|–||–||–||A||T||“Sorry, Wrong Number”||–||C||V||–||T||She Wore A Yellow Ribbon|
|N||–||–||–||T||They Live By Night||–||–||V||–||–||13 Rue Madeleine|
|N||–||–||A||T||The Window||–||–||–||–||T||Yellow Sky|
|–||–||–||–||–||Ace in the Hole||–||–||V||A||–||All About Eve|
|–||–||–||A||T||Angel Face||–||C||V||–||–||An America in Paris|
|N||–||–||A||T||The Asphalt Jungle||N||C||V||–||–||Cheaper by the Dozen|
|–||–||–||A||T||Enforcer||–||–||–||A||T||Day the Earth Stood Still|
|–||–||–||A||T||Gun Crazy||–||–||–||A||–||People Will Talk|
|N||–||–||A||–||In A Lonely Place||N||–||V||–||T||Red Badge of Courage|
|N||–||–||A||T||The Big Heat||N||C||–||–||T||Johnny Guitar|
|N||–||–||A||T||Kiss Me Deadly||–||–||V||–||–||Sabrina|
|N||–||–||–||T||Touch of Evil||–||–||–||–||–||Witness for the Prosecution|
The approach David Cornwell and I adopted to studying film noir may be contrasted with some characteristics of the existing literature of the field. This is particularly clear if one considers our analysis of openings. We have concluded that most films in the noir category have “tense “ beginnings. Our claim is open to being checked by others, but it seems to me that it is unlikely that our judgments would be so undermined as to show support for Vernet’s rival claim that films noirs display quietude as they begin. Vernet discusses only six films but it is not simply in quantity that we believe our approach is superior. We explained how our sample was created. Vernet (p. 9) says that they were chosen “by the frequency with which they are shown and by the pleasure which they give me’. The latter explanation hardly seems appropriate for scholarly discourse, but Vernet may here simply be acknowledging openly something which is usually unstated. A more profound objection to Vernet’s intellectual style is the bias in favour of theory building as opposed to considering evidence. This is most clear where he acknowledges that the opening of Double Indemnity which he discusses is not actually the opening. However, his model proposes a distinctive type of first sequence followed by a second sequence. With respect to Double Indemnity, he describes what appears on the screen as “a question of the inversion of a model which is typical of the other films, but whose final result is…identical to them“ (p.5). One might ask here why, if the final result is the same, there is any particular reason for the supposedly typical quiet opening.
Three of the variables we explored found films noirs not particularly different from their contemporary non-noir film. Thus basing a film on a novel and employing voice-over narration were quite common features of Hollywood films of that era. The result for colour reminds us that this “film-noir” period was when colour gradually emerged , so “black and white” is hardly a defining feature of noirs taken in their historical context. We may note too that some writers have a category “colour film noir”, by which they both exclude a colour film automatically from the category “film noir” and implicitly define a noir film as being shot in black and white.
Our consideration of voice-over lends some support to Sarah Kosloff’s suggestion that it was a common device in the 1940s. We found it in over one in five of our noir sample and almost twice as frequently in the non-noir films (see Table 4). Thus although it appears to be characteristic of 1940s films it is not especially associated with noir. However, turning to the films of the 1950s, we did not find support for Kazloff’s claim that the decade saw a decline in the use of voice-over. It should be noted that our investigation was not primarily aimed at testing that part of her thesis. We sampled only 20 films in the 1950s, as compared to 64 in the 1940s and, in order to match the relative decline in the number of noir films in the lists we analysed, our sample was biased towards the earlier part of the decade. It may be that a more detailed inspection of films from the late 1950s would support Kozloff’s position.
We found that half our noir sample films were based on a novel, a somewhat higher proportion than was the case for the non-noir films. This is certainly sufficient to justify further exploration of what we have seen termed “crime literature”. This might involve two lines of enquiry. One would be writing other than novels, such as short stories or radio plays. Sorry, Wrong Number, one of our noir sample, is an example of the latter. The other would be to consider the screenwriting work of exponents of “crime literature”, such as Raymond Chandler. His novels were the origins of three films in our noir sample, The Big Sleep, Murder, My Sweet, and Lady in the Lake. However, he also has screenwriting credits for The Blue Dahlia and Double Indemnity.
TABLE 4 VOICE-OVER
|Based on Novel||36||With Voice-over||28%|
The setting of a film in contemporary urban America turned out to be the most distinctive characteristic we looked at. Florence Jacobowitz (p. 152) quotes Fritz Lang as considering his film Scarlet Street as being “a realistic film of the people and for the people “ (but she cites no source). If indeed these films are dealing with life as experienced by the contemporary audiences, then we might have been better to distinguish between “settings” and “protagonists”. Some of the films which were not predominantly SET in cities in the United States, nevertheless had American in central roles in these contemporary dramas. Examples include the characters played respectively by Glenn Ford and Robert Mitchum in Gilda and Out of the Past.
One final point which needs to be made is that in considering these features of films scoring 10 to 12 on the CH index, we are not arguing that only these are true films noirs. It seems to me that it is more appropriate to consider not which films are and are not “noir”, but rather to look for characteristics which give rise to the label “noir”. Some films may have several such characteristics and others fewer. Thus there may be degrees of “noirishness”.
Jacobowitz, Florence (1992). The man’s melodrama: The Woman in the Window & Scarlet Street, pp 152-164 in Cameron, Ian (Ed) The Movie Book of Film Noir. London: Studio Vista.
Kozloff, Sarah (1988). Invisible storytellers: Voice-over narration in American fiction films. London: University of California Press.
Vincendeau, Ginette (1992). Noir is also a French word; The French antecedents of film noir, pp 49-58 in Cameron, Ian (Ed.) The Movie Book of Film Noir. London: Studio Vista.
Vernet, Marc (1983). The filmic transaction: On the openings of film noirs. The Velvet Light Trap, 20, 2-9.