Forecasting the Future Feature: How Film Industry Hierarchies Shaped Trailer Discourse, 1919–1959

The coming attraction film trailer has successfully maintained its prominent role within film promotion for over a hundred years. This article explores the shifting historical status of the trailer within the film industry and how industry trade press reported on its development and widespread adoption. Across this period these publications worked to delineate the discursive borders within which trailer debate occurred: from attacks on the trailer’s usefulness to related claims of accuracy and fidelity. Exploring the creation of this discourse challenges the idea that the increasingly negative tone around the film trailer in the twenty first century is a uniquely modern phenomenon. The article argues that these initial industry strategies need to be understood in relation to key cultural and industrial concerns around commerce and artistry, critical cultural gatekeeping, and broader interests in forecasting. By focusing on a largely overlooked element of the classical Hollywood system, we demonstrate how trailers existed in a disputed space within that system: a crucial promotional tool but also a creatively potent film text.

The coming attraction film trailer has successfully maintained its prominent role within film promotion for over a hundred years.This article explores the shifting historical status of the trailer within the film industry and how industry trade press reported on its development and widespread adoption.Across this period these publications worked to delineate the discursive borders within which trailer debate occurred: from attacks on the trailer's usefulness to related claims of accuracy and fidelity.Exploring the creation of this discourse challenges the idea that the increasingly negative tone around the film trailer in the twenty first century is a uniquely modern phenomenon.The article argues that these initial industry strategies need to be understood in relation to key cultural and industrial concerns around commerce and artistry, critical cultural gatekeeping, and broader interests in forecasting.By focusing on a largely overlooked element of the classical Hollywood system, we demonstrate how trailers existed in a disputed space within that system: a crucial promotional tool but also a creatively potent film text.
In 1959, Martin J. Quigley in the Motion Picture Herald hailed the 50 th anniversary of the foundation of the National Screen Service Corporation and celebrated the film trailers the company specialised in producing and distributing: Trailers, appearing on thousands of screens, before vast, receptive audiences proved the perfect medium for building fan enthusiasm, and compelling the public to return again and again … no other industry … has so potent a medium, so powerful a means of selling its product, at so small a cost, as the trailer … its high degree of effectiveness has been proved beyond all doubt by such imposing polls as those conducted by Sindlinger, the Opinion Research Corp. of Princeton and National Theatres' survey. 1 The coming attraction film trailer has now been part of the commercial film industry for over a hundred years.In that time the trailer has successfully weathered a series of challenges to its dominant role within film promotion: adapting to new technologies, new exhibition platforms, new aesthetic approaches, shifts in production priorities, and industrial contraction and expansion. 2Trailer scholarship over the last two decades has mainly focused on the textual content of, and potential spectatorial responses to, these coming attractions, with much less known about the historical status of the trailer within the film industry: what elements of trailer content were most prominently discussed or debated; and was the reception of its industrial expansion between 1919 and 1959 truly as positive and unchallenged as the Quigley quotation suggests?
In this article we focus on the introduction of the film trailer to the Hollywood and British film industries to assess the 'reception trajectory' of the trailer during its first fifty years: that is, before, during and after the main period of the classical Hollywood studio system. 3Our analysis will reveal how the film industry trade press, alongside other popular and fan publications, established a series of discursive markers that delineated the borders within which trailer debate occurred: most notably, attacks on the trailer's usefulness and related claims of accuracy and fidelity.From this evidence, we argue that these discursive strategies need to be understood in relation to key cultural and industrial concerns: the division between commerce and artistry, critical cultural gatekeeping, perceptions of the film audience, and broader cultural interests in forecasting.By positioning the trailer in relation to this discourse, we reveal the disputed space trailers were able to maintain within the classical Hollywood system: a crucial economic and promotional tool yet one that, by virtue of also being a creatively potent film text, challenged the centrality of the feature film experience.With historical emphasis often falling more on feature and non-fiction filmmaking (over the so-called 'useful' cinema), this case study of trailer discourse adds an important perspective to the classical studio period. 4By exploring the roots of industrial trailer discourse, we stress how any understanding of the more recent negative response to trailers must be rooted within that historical trajectory, not seen as an uniquely modern phenomenon.
The diachronic focus on historical film industry discourse that sits at the heart of our article is an expansion and response to existing studies of film trailers.As noted above, much of that scholarship has been based around close readings of film trailers that investigate topics as diverse as rhetoric, genre, fan production, authorship, technology, and psychoanalytic film theory. 5Within that, scholars such as Janet Staiger, Keith J. Hamel, Fred Greene, and Keith M. Johnston have begun to highlight aspects of the earliest years of trailer history: debates over what counts as the first trailer, the link between trailers and pre-existing promotional materials (such as the theatrical series slide, poster, or press book), and the slow emergence of a trailer format that we would recognise today. 6In particular, Staiger and Hamel both note how the trailer emerged at a time when the producer-distributor model was beginning to take hold in Hollywood as the industry shifted towards its eventual studio system form: that model pushed for more centralisation of activity across the 1910s and 1920s, including more control over the production of advertising materials that would ensure similarity across all exhibition sites. 7It was this desire for centralisation and control that fuelled the creation of National Screen Service Corporation (hereafter NSS) in 1919: described as a bid to get the nascent studios 'out of the nickel and dime business of selling trailers and posters and stills to individual theatres'. 8he 'advance strip of film' issued for The Quest of Life (Miller 1916) 'to give the public a foretaste of what the photoplay will provide' was one of several pre-1919 trailers featured in industry trade press.9 The trailer format saw exponential growth over four years: Moving Picture World commented in 1917 that 'the trailer … has come into use extensively of late' and by 1919 the major film producing companies had agreed to create a new company, NSS, to produce and distribute their movie trailers and other advertising accessories.10 NSS established its main office on Time Square in New York, near the studio's East Coast sales and marketing departments; it set up regional sales and distribution hubs across America through the next decade; and secured crucial business deals that embedded NSS trailer production staff within the major studios.11 Set up and run by ex-studio executives, designed to fall in step with the developing 'mature oligopoly' of the studio system, NSS monopolised the production and distribution of film trailers from 1919 through the late-1950s: mirroring the studios to such an extent it experienced its own government monopoly investigation in 1948.12 As we will detail below, trailer production and content remained the focus of most industry discourse, but NSS's real business lay in distribution.President Herman Robbins set up a system whereby individual theatres paid 'annual licence fees (from ten to twenty dollars per week) … for the right to use its [NSS] products': ensuring both NSS's dominance and the push towards wider studio control.13 While certain film studios such as MGM or Warner Bros. would occasionally produce their own trailers in-house, they continued to rely upon NSS's national distribution network.Given this dominance, our analysis below remains conscious that any trade industry discourse on the trailer is also a commentary on NSS, either in the US or the UK.Indeed, the British arm, National Screen Service Ltd., was set up in 1926 and created a similar monopoly in that national industry.14 The existing scholarship around trailer history has been extremely important in beginning to sketch out a picture of the first decade of trailer production, but a fuller understanding of trailer history remains elusive.This is, in part, due to the absence of a rich vein of primary archival production materials that are standard within recent film histories: for example, the lack of paper traces relating to the production and industrial activities of the British NSS required telling its story through a patchwork of other historical sources.15 In one of the first book-length studies of the trailer, Lisa Kernan identified an approach to assess the trailer industry that relied upon 'extratextual documents … [to] shed light on the industrial, institutional and cultural influences that shape both audiences' interpretations of films and the ideological underpinnings of Hollywood production practices.'16 Ultimately, Kernan did not choose this path, opting instead to focus on how trailer analysis might reveal 'the film production industry's assessment of its actual audience': but that potential alternate route helped fuel our approach here.17

Methodology
To undertake a discursive study that would allow us to offer a more forensic account of trailer history and reception, we were initially inspired by Janet's Staiger's 1990 article and her argument that 'asking and answering questions about the production of advertising does not answer questions about its reception'. 18nowing that information was already scarce on NSS production and more interested in tracking industrial discourse across a longer period we adopted and adapted terminology from Ernest Mathijs, specifically his concept of a reception trajectory: reception is rarely unitary … it consists of many competing and opposing discourses, struggling for dominance and stretched out over several time frames … What is needed is what I call a reception trajectory: an integrated view of specific discourses operating in particular situations (synchronically) and as processes over time (diachronically), all analysed as different types of "talk" about film … in order to map both the individual strategies used to forge meaning (or fail to), as well as their combinations … This provides a view of the complex pattern of influences and opinions that makes up a film's reception'. 19like Mathijs' work on Les lévres rouges (Harry K} umel, 1971), or other reception studies, our focus is not an individual film, film cycle, or genre, but a parallel industry that interrelates and overlaps with commercial film production.The key sites of industry discourse identified are articles and reports found in film industry publications such as Variety, Film Daily, Kinematograph Weekly, Motion Picture Herald, Motion Picture World, Picturegoer and Cine-Technician.The bulk of the material was accessed using the Media History Digital Library, an important and now ubiquitous source of primary materials from the film industry trade press. 20For all its wellearned place within digital film history, identifying and analysing a topic as broad as film trailers, within the time-period envisaged, presented its own challenges.A basic search for 'trailers' across all materials and time range returned over twentytwo thousand hits (n ¼ 22,098).To allow us to better ascertain patterns and themes from within the dataset rather than impose our own interpretative frames at the outset, we applied additional coding and analytical decisions to deal with the wealth of information.Narrowing the search to the period of NSS's monopoly and the classical studio system (1919-1959) reduced the overall number by three thousand (n ¼ 19,504).We then chose to sample the pre-1919 period to check options for narrowing our scope further while remaining confident we could capture relevant discursive elements for our study.
As described above, this period has limited extant examples of trailers and even the basic terminology of 'trailer' is not yet fixed. 21From the 686 records on Lantern (35 in 1910-1914, 625 in 1915-1919), we sampled ten per cent of those records (n ¼ 68): this revealed a range of articles and examples where short 'trailers' were made for non-advertising purposes such as censorship announcements; where 'trailer' was used as a term for lagging behind competitors; or where short propaganda films promoted topics such as diverse as coal conservation or Red Cross campaigns. 22These results support Janet Staiger's claim that the term 'trailer' had a wider application in this period while popular and industry definitions began to coalesce around the short 'coming attraction' format that would come to dominate. 23The potentially shifting nature of the term did represent an important point of consideration regarding our corpus, not least as a reminder that such broader definitions might recur in the wider sample.
Applying what we learned from that small sample to the main selection of 1919-1959 records (n ¼ 19,505) allowed us to identify a series of trade press articles, fan magazine commentary, and advertisements (often for NSS, its competitors, or studio-produced trailers).By covering a diverse range of the available publications we hoped to capture different views (or different interpretations of shared views); equally, we chose not to aim for commensurate coverage of each of the fifty years, instead ensuring that our sample offered similar coverage across each decade.To this end, the full survey of articles we undertook (n ¼ 1950) offered the following breakdown: 1919-1929 (n ¼ 435); 1930-1940 (n ¼ 660); 1940-1950 (n ¼ 450); 1950-1959: n ¼ 405).Articles were analysed and coded depending on high numerical repetition of topics with direct relevance to industry commentary on trailers.This meant that from the overall sample we put aside those reports that covered industrial topics not wholly relevant to industry discourse or commentary on trailers, for example: movement of staff within companies; reports on annual N.S.S. conventions; details of industry deals or agreements; or, in relation to the broader nomenclature of 'trailer' discussed above, notes on studio-or government-led schemes that had adopted trailer-style productions to celebrate seasonal holidays, sell war bonds or promote charitable causes. 24f the articles that remained, different areas of industrial 'talk' emerged: around fixing the trailer, trying to define what a 'good' trailer should do, and what restrictions or approaches might achieve that.Additionally, key discursive terms such as 'spoiler', 'misleading' (or 'accuracy'), 'inappropriate', and 'quantity' spanned the whole diachronic trajectory, with some synchronic variation (detailed below).We noted that the trade press commentators whose voices can be found across this discourse saw themselves as speaking for the audience or viewer who is unable to assert themselves in the same way.As such, our analysis posits that the reception of trailers in this period should be seen as an act of gatekeeping, with commentary working to establish and reinforce historically contingent views on the film trailer.
Given the limitations of our sampling method, our analysis cannot offer a comprehensive survey of fifty years of industry responses.What our data analysis has revealed are key discursive tracks that suggests a dominant reception trajectory: shifting and overlapping patterns of critique and dismissal that speak to the disputed industrial value of the trailer from the 1910s through the late 1950s.As detailed below, this discursive trajectory privileges certain hierarchies and areas within the film industry: most notably the voices of exhibitors and press commentators that claim to speak for both the industry and the cinema audience themselves.Yet our data also features individual audience members, appearing through letter columns in movie magazines such as Picturegoer: these claims are often in concert with industry commentary, but also need to be understood in terms of the wider ideological perspectives of those publications, their industrial role, and their ability to control which letters were published. 25elow, we consider three dominant areas of the trajectory in depth and note how they map onto identifiable issues around gatekeeping, debates around creativity and commerce, and interests in forecasting and salesmanship.Teasing apart these discursive strands reveals shifting and overlapping perspectives that speak to the growing pains of the trailer industry and remain suggestive precedents for trailer discourse in the modern industry.

Tell-Tale trailers: the emergence of spoiler discourse
"Trailers" of pictures, flashed on the screens of our local movie emporiums the week before the entire picture is shown, are in my opinion "picture spoilers," instead of well-planned advertising that the exhibitors intend them to be. 26e accusation that trailers contain spoilers is not a uniquely twenty first century phenomenon, with trailers being accused of spoiling movies for almost a century.In April 1926, when trailers were still very much in their infancy, Variety reported on the broad 'variance of opinion among exhibitors as to the advisability of too much action being shown by the producers in the trailers … some say that some of the trailers show some of the big scenes and that they take away the edge when the picture comes to the house.' 27 Over the next decade, Variety continued to report on spoilers in trailers, noting their overuse had led some to claim it 'hurts biz and have eliminated those with thrill scenes'; and a continued exhibitor dissatisfaction over trailers showing 'too much' which was linked to claims of audience unhappiness. 28Trade journal ads for NSS' short-lived competitor, Exhibitors Screen Service (hereafter ESS) presented themselves as the cure for 'scene trailers [which] … disclose plots, thrills, high spots'.ESS trailers claimed, in a thinly veiled riposte to the monopolistic NSS, to be separate from such hyperbole, creating trailers that do not disclose 'important scenes and sequences'. 29 New York Times report on the growth of movie theater advertising in the 1930s noted trailers were a form of advertising which 'occasionally [provokes] audience resentment' because they 'sometimes give practically the entire story'. 30 fan voicing displeasure in the pages of Photoplay found fault with early trailers because they can 'completely relieve a film of its important thrill of suspense'. 31icturegoer regularly published fan complaints against the trailer and its perceived tendency to spoil or kill suspense in the film advertised.A 1938 fan letter asked, rhetorically, '[wouldn't] you feel rather tired of seeing the big moments of the film so many times in advance and lose your inclination to see the show?I know I often do!' before then asking the Picturegoer editors to 'agitate for the abolition of trailers' on this basis. 32Less revolutionary correspondents merely vented their frustrations with trailers showing all the best scenes, gags, or climaxes. 33any of these complaints raise the quality of trailers in comparison to the films they're supposed to sell.In other words, the trailer appeared to work too well and was arguably better than the film: thus superseding its proper place in the hierarchy.Tom Matthews highlighted a 1926 Los Angeles Times report that people were clapping a trailer, 'express[ing] as much enthusiasm for it as they did in the feature presentation'. 34Other correspondents and viewers had similar views: The Trailer shows you all the high lights and exciting scenes of the film and, true enough, the scenes shown in the trailer are excellent.They are so good that you do come back to see the picture it advertises, but to your disappointment, you find that the Trailer has been more than generous.It gave you all the important scenes of the picture … The result is that after this happens to you two or three times, you don't go back to see the film. 35milarly, a Fox-affiliated exhibitor writing in the Fox West Coast Theaters Now company newsletter described how frequently his patrons 'tell me they like the trailer better than the picture'. 36This exhibitor then recommends dialling trailers back so that they don't show many scenes and thereby give too much away to audiences: 'A few scenes might be shown, but not all the punch scenes … sell without showing them'. 37uch complaints against the trailer, its potential as a source of audience pleasure, but also its potential for spoiling the future pleasure of the feature film were often coupled with suggestions on how to fix them, create a new form of trailer, or simply how to cope with them.The underlining principle in each worked to ensure that a trailer would remain in an appropriate (and subservient) role to the feature it was designed to sell.This can be seen as early as the 1920s, with reports on producers experimenting with a spoiler-free style of trailer, 'animated with title, cast and pictures of the principles in character make-up, all scenes being eliminated'. 38As NSS had secured contractual access to feature film footage from the major studios these new trailers were often promoted as spoiler-free by companies such as ESS (or the British-based firm WinAds) in an attempt to sell their own wares.This was claimed to be a response to those viewers who wanted trailers pared down to simple announcements and endorsed by Picturegoer which noted 'to show the cast is quite enough to draw the public to-day'. 39poilers remained a critique across the fifty years analysed.Each new iteration of this debate saw the potential for trailers to spoil a film as unique to that historical period, reliably and repeatedly presented as a threat to a proper, hierarchical relationship of film para-and primary texts.Trailers can spoil and thereby disrupt, supersede, or render pointless the films they're trying to sell.In this way, the recurring desire to fix trailers by eliminating spoilers suggests another desire on the part of the industry: to discipline the unruly qualities and possible disruptions caused by the creative aspects of a trailer into an ideal textual, industrial, and economic relationship based around accuracy.

Fidelity and forecasting: the misleading trailer
In the cloud-cuckoo trailerland where the adventure is never less than unparalleled, the Technicolor always dazzling, the spectacle regularly the most stirring we have ever seen, and the romance unfailingly breathtaking, the picturegoer is a familiar and unimpressed visitor … To cut down on the ludicrous over-exaggeration, though, is not enough.To fulfil its function completely the trailer should make some effort to capture the individual flavour of the film. 40is 1950s critique contains much that is familiar in repeated claims around the trailer as a misleading format: a stress on hyperbole, a beleaguered viewer, and a call for more accuracy and individual focus.Following on from a discussion of the historical nature of the spoiler discourse, this section analyses a series of articles where the trajectory of industry talk relates to claims that trailers mislead an audience; with a related demand for increased fidelity between the trailer and the feature film.As a debate that has, in the twenty first century, become a source of legal challenges, it may be useful to investigate where this strand of the discourse originated. 41n 1921, a commentator in Exhibitors Trade Review claimed exhibitors appreciated 'the artistic and mechanical efficiency' of a trailer. 42Five years later, Shirley S. Simpson was still able to claim that a good trailer offered a 'comprehensive and definite idea of the quality of the film'. 43By the end of the decade, the tone had shifted: it seems as though the writers would run out of superlatives … patrons are always promised the finest entertainment ever made … Perhaps these "trailers" might really arouse the spectator's interest if they were set forth in a more judicious fashion, with a conservative wording and more rational and less sensational selection of the excerpts from the films. 44though the trailer is not specifically tagged as misleading here, the key request for a more 'rational and less sensational' use of scenes presages the focus on accuracy that would be honed and elaborated over the next decade: echoes of that sensationalist claim can be seen in articles linking misleading trailers with a tendency towards exaggeration and hyperbole that affects narrative and generic accuracy.In 1932 a Picturegoer reader described filmgoers being 'spoofed into a false valuation by those artfully selected excerpts'. 45In the same year Film Daily posed the question, 'Is present-day theater advertising too inclined to exaggerate?' claiming there was an almost unanimous agreement around 'the advisability and advantages of making ads more informative and less misleading'. 46Key terms that emerge from individual critic responses include 'exaggerated ballyhoo', claims of misrepresentation, and 'broken promises' that test 'the public's faith'. 47In each case, critical talk centres on the idea that the trailer offers an inaccurate representation of the feature film.This moment, and critique, is best encapsulated in Howard T. Lewis' 1933 book The Motion Picture Industry: it must be said that many trailers are both misleading and ineffective … they do not fairly represent the real character of the play.In other words, though the trailer may be actually composed of sequences from a picture, the picture itself is not fairly sampled, but is actually misrepresented by the trailer. 48e 1930s is the decade when the borders of this critique around selection of material are most clearly set out.Trailers are misleading because they do not represent the feature film.Trailers use too many scenes that are not well sampled or chosen.Trailers are too sensational and hyperbolic compared to the feature film.This is a claim that echoes down the years, both in terms of popular discourse, but also academic work around the paratext. 49But what underpins such critical reception talk that sees the trailer as a misleading text?
The hierarchies revealed across our analysis of this critical and industrial discourse represent a form of cultural gatekeeping: the trailer should not detract from, or subvert, the primary entertainment form of the feature film; articles shared a desire to define what footage was considered appropriate for an audience to consume; and commentators felt the changes they demanded were given in the name of protecting that audience.To such critics, those 1920s claims of trailer artistry or mechanical efficiency had been proved false, possibly through repetition and familiarity.The trailer was accused of having shifted into a more hyperbolic mode.It was not offering an effective sample, therefore it was not performing its financial role.Yet, as seen in the previous section, it had also created itself as a popular form of creative entertainment that regularly ran the risk of outshining the feature.
The gatekeeping role adopted by industry commentators discursively pivots between those binaries of art and commerce (or artistry and efficiency): the trailer sits awkwardly in the middle, using the material of art but reworked into a commercial format, with the specific instrumental goal of increasing industry revenue.As Lisa Kernan has noted, the trailer remains 'a unique form … wherein promotional discourse and narrative pleasure are conjoined (whether happily or not)'. 50he repeated complaint in our dataset that the trailer is not an accurate representation of the feature is a statement that the trailer's reworking (or adaptation) of the text somehow reduces or undermines its cultural or artistic value because it lacks fidelity.Yet the debate also swings back to commerce, with claims of assessing accuracy linked to a desire to protect the viewer -the bearer of the ticket, the decider of a film's financial success or failure -from the worst excesses of this capitalist system.
The American cultural context within which the trailer was created might offer some explanation for this recurring discursive emphasis on accuracy.In the late nineteenth and early twentieth century American society had a specific obsession with scientific claims that it was possible to forecast the future: The modern forecasting field … had many points of origin in the previous century; in the credit rating agencies, in the financial press, and in the blossoming fields of science -including meteorology, thermodynamics, and physics … But forecasting also had deeper routes in the inherent wish of human beings to find certainty in life by knowing the future … forecasting addresses not just business issues but the deep-seated human wish to divine the future.It is the story of the near universal compulsion to avoid ambiguity and doubt and the refusal of the realities of life to satisfy that impulse. 51e rise of scientific methods that claimed to reveal patterns or future trends within the economy presages much of the industrial and cultural worlds within which the trailer emerged.As Friedman notes, the very word 'forecast' combines something that comes in front or in advance (fore), with a calculation or estimation (cast): 'forecasting means to calculate the future before it happens.' 52The adoption of the word trailer as the industrial term for the 'advance strip of film' can be seen as a fulfilment of that cultural desire for knowing the future, be that economic, personal or, in the case of the coming attraction, entertainment. 53hile it may be an artistic form in its own right, the trailer is specifically offered to an audience as an industrial forecast of future pleasure, a calculated estimation of what is yet to appear.In the case of the trailer, such estimates were usually based more on concrete production knowledge than those individuals who promised millions that they could forecast the stock market.As such, the trailer is held to retrospective critique when the forecast appears to be false, mirroring economic forecasting where 'customers … did not seek out detailed information on the accuracy of economic predictions, as long as forecasters proved to be right at least a portion of the time.' 54 The longevity of the trailer may be located in that contradiction: the desire for knowledge even if, on occasion, that knowledge does not achieve the level of accuracy required by each individual.Yet providing the trailer matches expectations often enough, its forecasting role remains safe.
That focus on forecasting speaks to a growing awareness, and adoption, of scientific methods across different areas of American industry; methods that were applied not only to forecasting, but also to production and salesmanship.Yet it appears that, despite claims being made for the positive or negative impact of the trailer on an audience, and its position between art and commerce, the Hollywood film industry was initially resistance to the development of high-level industrial and consumer research that was fuelling shifts in how corporations understood sales management. 55While industry market research on trailers and other forms of promotion is now more significant, robust, and widespread than in the first half of the twentieth century, the results of this research have historically remained largely invisible to audiences as well as industry commentators and professionals.In its place, press commentators, much like studio executives, appear to have often relied on hearsay, fan mail, or their own subjective feelings about the trailer.As Leo Handel points out, 'we hear that movie making is basically an artistic endeavor.We would gladly accept this statement if the same people did not tell us, after turning out a series of utterly commercial cliche pictures without batting a solitary eyelash, that movie making is just a business like any other'. 56The industry talk that grew around the trailer, then, is best understood as an attempt to control this most visible and emblematic form of the commercial movie industry, a form of pleasurable and narrative-based forecasting that was also required to perform an economic job.The 1930s may be the point this discursive element becomes foregrounded within our reception trajectory but following decades developed aspects of it.In the 1940s, studios were asked to 'engage trailer producers whose principal qualities are a keen sense of proportion coupled with restraint'. 57Equally, 1950s commentators saw a return to 'bad and misleading film trailers', a suggestion that they were already nostalgic for a previous generation of trailer production. 58The idea of keeping faith with the public remained central, although little evidence was offered that audiences were put off by the trailers being critiqued: in 1935 Picturegoer published a letter complaining about 'terrible trailers … [that] give a very erroneous impression of coming films … usually boosting a film high above its merits' 59 ; fifteen years later another letter in the same publication espoused that trailers 'should be abolished' because as 'guides to the merits of the films they purport to advertise, they are valueless … More often than not they are grossly misleading'. 60The similarity across that period (and the wider dataset) underpins the idea that trailers' ability to effectively forecast the future was always balanced against the retrospective assessment of each individual viewer.
Despite the recurring nature of this element, the same 1950s trade press publications would also foreground reports that trumpeted the trailer industry as an exemplary part of the American and British film industries.Alongside the critical commentary summarised above, these magazines continued to report on scientific surveys where trailer effectiveness was praised: '342 out of every 1000 people attending the theatres surveyed were motivated primarily by the trailer to return for the next attraction … Almost 85 per cent could 'play back' something they remembered from the trailer they had seen'. 61There was no evidence in our data that these publications offered any link between such positive reporting and the clear prevalence of negative commentary, a suggestion of the overlapping discourse that could appear even within the same publications.
NSS appears to have been aware of this ongoing industrial debate around accuracy.The strategies they adopted to tackle it looked not towards the audience but their licence fee-paying clients in exhibition or their contractual partners in the studios.From the late 1920s on NSS advertising campaigns in the US christened the company the 'Prize Baby' of the industry, most often illustrated with an image of a baby (in different outfits such as a nappy or a top hat and tails) with text that hailed exhibitor success, box office returns, and claims about the success of trailers in attracting audiences.Such advertising was exclusively connected to the commercial impetus within the trailer industry rather than any claims of creativity.By foregrounding techniques that stressed sales and success, NSS was borrowing from shifts in large companies such as Eastman Kodak and Westinghouse who had demonstrated 'how old traditions of selling, persuasion, prediction, and motivation were brought into the rationalised world of managerial capitalism'. 62Within our sample, the recurrence of adverts citing trailer statistics and reports on the activities of regional sales offices reiterate how NSS had responded to the new culture of salesmanship within American companies: 'a strong sales department was essential for the success of these large firms, helping them to generate demand and to prevent competitors from entering the industry … To use a military analogy common in the early twentieth century, advertising was a weapon for waging an air war, while salesmen were deployed as foot soldiers in a ground campaign'. 63By promoting their business at a national level, while expanding and supporting their customer-facing regional sales forces through sales initiatives, annual conventions and a company magazine, NSS clearly hoped to imitate the success of those other corporations.
One final example from our sample encapsulates the difficulty of parsing the critical discourse around accuracy while acknowledging the synchronic nature of some aspects of it.A late 1940s exhibitor survey made a series of recommendations designed to improve the trailer.Targeting the claim the trailers were too misleading, the advice was to 'minimise superlatives … not reveal too much of the plot or too many of the best gags … [and] not use critics' comments'. 64hile broadly generic, these clearly speak to a desire for the trailer to offer a better sense of the feature film, stressing its potential to give a more accurate forecast of the future film experience.Other recommendations, however, appear to actively advocate for trailers to mislead, through the deliberate exclusion of key features: 10.Trailers should not publicise directors of pictures 11.Trailers should avoid use of costumes wherever possible 12.Trailers for English pictures should use American commentary 13.Trailers dealing with controversial subjects should emphasise entertainment and not the preachment. 65ven the sources of the recommendations, these identify key exhibitor beliefs: that a director might not sell a picture; that costume drama is off-putting; disguising a British film by using an American voiceover might make it more palatable; while avoiding a difficult topic might lure an audience in.These all speak to a clear desire for the trailer to elide or actively subvert the specific features of the film that exhibitors did not like.Or, to put it another way, in listing ways to improve the accuracy of a trailer forecast, these 1940s exhibitors were actively advocating for the trailer to wilfully mislead the audience.
The contradictory nature of this debate, not least its complex relation to cultural fascination with forecasting, claims of audience response, and different ideas of how best to assess accuracy, means it was rarely absent from trailer discourse.As with the spoiler debate, individual decades suggest some synchronic differences: but the overall themes of gatekeeping and tensions around creativity and economics remain dominant through 1959.Given the prevalence and accessibility of trailers in the twenty-first century, it is perhaps not surprising this element of industry talk has recurred with a renewed ferocity: after all, at no other time has it been so easy to assess the accuracy of a trailer's forecast when it is available digitally for multiple repeat viewings.

Too many trailers
Spoilers and fidelity are two sides of industry talk around accuracy: in one, the trailer is too accurate and revelatory in its forecast, in the other, the forecast is retrospectively read as faulty or misleading.Even when other aspects of industry talk were present (such as censorship or claims of inappropriate content) most were underpinned by notions of an accurate forecast.A smaller, although still significant, grouping of data sits at a similar intersection between studio, exhibitor and audience: where the trailer is regarded as overstaying its welcome either in terms of individual running time or in an increased slot within film programmes.Industry talk around trailer length and number begins almost in parallel with the introduction of the studio system and trailer industry.In 1917, Motion Picture News correspondent W.S. Bach claimed that 'the new stunt of using trailers' in cinemas had snowballed to such an extent that theatre managers had 'a few more than … really needed' but would put them into the programme anyway.Citing a recent cinema visit, Bach continues: one little trailer started showing Mary Onderdonk in about a hundred feet of her newest film that was to play the house next Monday -then another piece of trailer showing handsome Bob Oklahoma saving his leading lady in a new picture, to show Tuesday -and so on and on … Dandy little trailers they were individually, but taken together they made a most beautiful mess … The little family by my side were disgusted … "I didn't pay to see an Irish stew of film about what's coming -I want to see the picture that showing tonight … why can't the fellow that is running this show realise that I didn't used to mind one of these half portion films, in fact I rather liked it, but six or seven of them gets my goat … these moving picture fellows always carry a good thing too far." 66ch of Bach's approach is familiar from other articles: the use of an audience example to root the complaint; an acknowledgement that trailers have their place; but ending with a broad condemnation of the industry that stresses Bach's own hierarchical position.Bach did not remain a lone voice for long, as 1930s commentators picked up this thread: 'too many of them are shown.I have seen programs here carrying five or six de luxe trailers'; '[theaters] are showing too many trailers'; 'be extremely careful not to give them too many trailers … no more than two or three … ought to be shown'; 'ten minutes of every performance [are] wasted by the showing of a trailer'. 67espite this, the 1948 exhibitor survey discussed above did not call for a reduced number of trailers but recommended that 'trailers should be shortened' and that a 'special short trailer is needed for multiple change houses'. 68Given this survey was of the exhibitors who were directly responsible for selecting how many trailers to screen, and who paid their licence fee to NSS, it may suggest they were less likely to criticise their own practices of trailer programming and preferred to see fault on the production side of the industry.The next obvious increase in trailer output into the 1950s was the lengthening of trailer duration, most famously for films such as The Ten Commandments (Cecil B. DeMille, 1956: tenminutes), or Ben Hur (William Wyler, 1959: four-minutes).The absence of negative responses around these examples suggests that the link to the epic blockbuster film was enough to justify an equally epic trailer: although it is not clear how such expanded trailers affected industry talk about too many trailers in the cinema.

Conclusion
In June 1926, around a decade after the trailer had become a regular promotional format, a National Screen Service advertisement claimed that its trailers were 'carefully produced … to whet the appetite of patrons, without giving away the story'. 69Even at this early stage the dominant trailer producer had clearly identified two crucial elements: the trailer was engaged in forecasting; and it had the potential to be over-revelatory.Here, as in the different threads of the historical discourse identified above, two underlying themes are clear: trailers need to be careful and restrained in what they choose to portray; and the trailer might be too good, or too effective, offering a coherent and enjoyable cinema experience in 120 sec rather than 120 min.The repeated attempts to corral, restrict, and limit the trailer -show less, be more accurate, don't spoil, be quieter, only teasepoint to the trailer's potent disruptive power for the industry which begat it.
The growth of the trailer from the 1910s through the classical studio era challenged the dominant mode of textual consumption that had emerged within the film industry, suggesting alternative pleasures to the intended creation of desire and anticipation implicit in the good trailer: that is, one that does not misrepresent the feature film.It is perhaps not surprising, then, that the industry talk that accompanied the emergence of the trailer saw the format as both a boon and a problem to the Hollywood studio oligopoly of the 1920s.At a time when the feature film became the industry norm, the trailer offered an alternative focus of audience attention, desire and pleasure.The financial role of the trailer can be celebrated because when it successfully promotes the feature, it remains a good object: yet when its content became troublesome, debate shaped it as a bad object in need of discipline and reshaping.
This article has outlined the concerted attempts by industry commentators, exhibitors, and occasionally audiences, to domesticate and control the trailer.The boundaries within which the good trailer should operate are repeatedly stressed: promote the film; give an accurate forecast; don't overstep creatively by offering or revealing too much.It has revealed tensions around how the trailer can be celebrated, often simultaneously, for its autonomous creativity and its contribution to the commercial success of the forthcoming media text. 70On the basis of these sources the evidence appears compelling but, in reflecting upon our methodological approach, we are reminded of the shifting linguistic and industrial definitions and uses of 'trailer' in the 1910s and after.Whether we came close to offering the 'transparent, unbiased process' that has been called for in studies of historical or contemporary trailers is for our readers to decide. 71However, we would note that at our 1959 endpoint the trailer had already expanded out to radio and television, a potential challenge to the film-centric nature of its dominant industry definition.
The discourses found across these fifty years stressed that the trailer should always perform a peripheral role, subservient and subsidiary to the feature film, ignoring the range of consumption possibilities an audience might bring to a trailer; and in so doing, sets and restates appropriate borders for the trailer.For example, the specially shot trailer has, at different times in trailer history, been derided for not showing clips from the film (therefore lacking accuracy). 72Yet specially shot trailers associated with an auteurist or showman figure such as Cecil B. De Mille or Alfred Hitchcock are allowed to temporarily move into a primary textual position through the involvement of a key creative figure sponsoring and controlling its transgressions. 73Without the rare endorsement and public presentation of such authorial figures, the trailer remained subject to a series of recurring disciplinary discourses, expressed by cultural gatekeepers within industry trade talk; all of which try to reign in the potential claimed excesses of the trailer in order not to upset the existing, and idealized, exhibition and viewing hierarchies.