Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
Imagined audiences: intuitive and technical knowledge in Hollywood
(USC Thesis Other)
Imagined audiences: intuitive and technical knowledge in Hollywood
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
IMAGINED AUDIENCES: INTUITIVE AND TECHNICAL KNOWLEDGE IN HOLLYWOOD by Stephen Zafirau ________________________________________________ A Dissertation Presented to the FACULTY OF THE GRADUATE SCHOOL UNIVERSITY OF SOUTHERN CALIFORNIA In Partial Fulfillment of the Requirements for the Degree DOCTOR OF PHILOSOPHY (SOCIOLOGY) May 2009 Copyright 2009 Stephen Zafirau ii Dedication To my parents, Samuel James and Catherine Zafirau, for their generous and encouraging spirit. iii Acknowledgements Writing a dissertation is mostly a solitary undertaking, but numerous people made this research project possible—and made the process of doing research more interesting and fruitful. I wish to thank my dissertation chair, Nina Eliasoph, for all that she did to help me craft the dissertation that I ultimately wrote. The creativity that she brought with her as she advised me on this project helped make the experience vastly more interesting than it otherwise would have been. My other committee members also provided crucial support as I went about doing this research project. Paul Lichterman left an important mark on this dissertation, particularly through the methodological training I received from him. Peter Clarke gave valuable and prompt suggestions on chapter drafts, and was also extremely helpful in connecting me with important contacts in the entertainment industry. I was truly privilaged to have him serve as a committee member while working on this dissertation. Others at USC also helped facilitate my research. Larry Auerbach, the Dean of Student-Industry Relations in the School of Cinematic Arts, gave me contact information for people working at high levels in the business of Hollywood. Without his willingness to help me with this project, many of the interviews discussed in this dissertation would never have taken place. I am also grateful to the Reference Department at Doheny Memorial Library. They were extremely helpful in allowing me to locate archival materials, and helped make sitting in the library’s basement and looking at old microfiche quite fascinating. The College of Letters, Arts & Sciences and the Department of Sociology at USC both, in their own ways, provided generous financial support for my research. The iv College granted me a Final Year Dissertation Fellowship, which was invaluable as I finished writing the dissertation. The Department of Sociology gave me small grants that financed a survey of movie executives as well as my attendance at entertainment industry conventions, both of which led to important insights. But, it is impossible to speak of insights I gained while doing this research project without acknowledging the help of those in the entertainment industry who appear in this study. Many were extraordinarily giving with their time in a business where time is extremely valuable. Although I cannot thank them by name, I appreciate the patience that many displayed as they attempted to teach me how the film industry worked. Last but not least, Sarah Stohlman was my partner in life as I worked on this dissertation, but she also felt like a full partner in my work. I cannot thank her enough for thinking about my research with me over countless conversations, reading more drafts of various papers than she ever should have been asked to read, and being my “Hollywood fashion consultant” when I so desperately needed it. I am indebted to her for the many happy memories I accumulated while in graduate school. v Table of Contents Dedication ii Acknowledgements iii List of Tables vi List of Figures vii Abstract viii Chapter 1: Finding Audiences in Hollywood 1 Chapter 2: The Evolution of Audience Knowledge in Hollywood 41 Chapter 3: Intuitive Knowledge in Reproduction 82 Chapter 4: Technical Knowledge in Reproduction 155 Chapter 5: Hollywood, Movie Audiences, and 229 the American Public Sphere Bibliography 234 Appendix A: Studying Hollywood Moviemaking 244 vi List of Tables Table 1: Institutionalized Audience Discourses and the Organizational 38 Contexts in Which they are Reproduced Table 2: A Sample of “The Grid” used in Script Coverage 90 Table 3: Survey of Movie Executives, Loglines for Hypothetical Movie 99 Ideas Table 4: Survey of Movie Executives, Ratings of Hypothetical Movie 100 Ideas: Mean Scores and Standard Deviations Table 5: Institutions of Reputation in the Hollywood Talent Industry 116 Table 6: Reasons Given in Limelight’s Script Coverage Summary for 146 “Passing” on Screenplays Table 7: Some Items Typically appearing on Test Screening 189 Questionnaires vii List of Figures Figure 1: The Circulation of Audience Knowledge in the Hollywood 35 System: Major Players and Processes Figure 2: “Optimism is Sweeping the Industry” 58 Figure 3: The “Lost Audience” 62 Figure 4: Understandings of Seasonal Box Office Returns, c. 1970 65 Figure 5: Variety’s Box Office Barometer, c. 1983 66 Figure 6: Variety’s Box Office Barometer, c. 1989 66 Figure 7: “Film Buffs” Advertisement, c. 1970 73 Figure 8: Viva Max! Advertisement, c. 1970 74 Figure 9: 1975’s Blockbuster Summer 78 Figure 10: Sculpture Near the Entrance of the International Creative 118 Management building in Beverly Hills. Figure 11: Sample Demographic Profile 218 viii Abstract Many scholars of the mass media highlight how the production of popular culture in contemporary media industries powerfully involves notions of “the audience.” That is, media producers draw upon collective assumptions surrounding who audiences are and what sorts of content they prefer as they go about making media content. This study proposes and develops an “evaluation in interaction” approach to examine how moviemakers socially construct American movie audiences within the context of their everyday work in the Hollywood film industry. Specifically, I show how moviemakers construct audiences using two institutionalized discourses of evaluation—both “intuitive” and “technical”—that are prevalent in the Hollywood film industry. This dissertation uses archival data to document the emergence and institutionalization of both intuitive and technical discourses of audience appeal over the latter half of the twentieth century. Then, drawing on organizational ethnographies of both a Hollywood talent firm and an audience research firm, as well as interview data with film industry professionals, I show how intuitive and technical discourses of audience appeal are reproduced within everyday work settings of the Hollywood film industry. Centrally, I argue that both intuitive and technical discourses about American motion picture audiences are embedded within the particular organizational conditions in which they are reproduced. 1 Chapter 1: Finding Audiences in Hollywood We are spending a billion dollars a year as an industry throwing wet paper towels at the ceiling. And when they by chance stick, we try to figure out what it was. Did we ball it correctly? Was it wet enough? Was the temperature of the room right? We don’t know…. And that’s why you see this herd mentality, where if something works, we all think it was because of the particular thing that got thrown. -Gil Cohen, 1 former studio executive, in a class on the movie business “No one ever sets out to make a bad movie.” I have heard this statement made on a number of occasions by those working at high levels in the Hollywood film industry. Yet, at the same time, these same individuals will concede that “bad” movies do get made. The apparent contradiction in these statements begs the question: what accounts for the frequent gap between cultural producers’ perceptions of a film’s commercial potential and that film’s actual performance? Or, put differently, how do those who make the movies think they are coming up with ideas that will be interesting enough to the 187 million Americans who buy movie tickets each year, and the many more who watch Hollywood movies outside theaters? As Gil Cohen’s analogy of paper towels on ceilings suggests, the question of what will “work” with audiences continually haunts the work of moviemakers. Every year in the film industry, thousands of potential movie ideas are talked about in the Hollywood film industry. They are pitched by screenwriters and discussed by talent agents, movie producers, and movie stars. Through multiple iterations, they are evaluated by executives and other studio personnel. And, they are studied and analyzed by professional audience researchers. Most of these ideas reach a 1 Any names of individuals, organizations, and other identifying information mentioned in this dissertation have been changed to protect the confidentiality of those observed. 2 dead end very early on in the process, and ultimately only approximately 200 feature length films are released in theaters during any given year by the major motion picture studios. Through all of the effort of moviemakers in considering these thousands of ideas that might potentially be a viable movie, how do they assess whether a movie idea will appeal to audiences? And when movies do seem to work, to what do moviemakers attribute this success? Despite the second-guessing that Gil Cohen and other moviemakers describe, it would be wholly inaccurate to characterize the process of evaluating movie ideas as anarchy. For the moviemakers I studied, understanding how and why movies “work” is much more patterned than is sometimes appreciated by some media scholars or moviemakers themselves. This is because understandings of movie ideas are organized by larger conventions within the film industry for how movie ideas might be talked about and evaluated. Moviemakers share collective langauages of audience appeal that are spoken in the disperse settings in which movies are produced, languages that cultural sociologists would call “institutionalized discourses.” A number of studies of the culture industries have highlighted the importance of these institutionalized discourses for talking about audiences or products as important ways in which decisions are framed within mass media organizations (see, for example, Ahlkvist & Faulkner 2002; Bielby & Bielby 1994). Such studies usually see these institutionalized discourses as being spawned amid conditions of deep uncertainty about whether or not products will “work” in the marketplace. In their study of network television development, for example, Bielby & Bielby (1994) concluded that in a context in which “commerical and critical success cannot be predicted in advance,” network producers use “linguistic framing devices of 3 reputation, imitation, and genre to reassure commercial and creative constitutencies that their actions are appropriate, legitimate, and rational” (1310). According to this interpretation, institutionalized discourses that are spoken within the culture industries (such as those having to do with reputation, imitation, and genre) are rooted in circumstances in which “nobody knows anything” about which products will work. Here, institutionzalized discourses provide legitimacy for products that, at their core, could only be judged idiosyncratically. What I argue here represents a departure from these studies of institutionalized discourse within the culture industries. While I do see these institutionalized conventions as important in structuring how producers talk within the culture industries, I do not think that these institutions stem from deep uncertainties about which products will “work.” Instead, what I attempt to show here is that institutions for talking about movie audiences in the Hollywood motion picture industry stem from the uncertainties that surround relationships within various sectors of the motion picture industry. Thus, from my firsthand observations of Hollywood moviemakers at work, I stress the importance of internal organizational conditions—in particular, field-level interactional institutions that govern how interpersonal legitimacy is obtained and maintained—in reproducing institutionalized conceptions of motion picture audiences. Put differently, I argue that moviemakers’ institutions for understanding and talking about audiences are made possible by the institutions for legitimate interaction that organize particular organizational fields within Hollywood film industry. Thus, institutionalized discourses about audiences not only help legitimate products as they are being conceived and 4 produced, but they also help individual people and firms to maintain their interpersonal reputations within their business networks. Second, I emphasize how the central institutionalized discourses in the culture industries deal specifically with evaluation. The commercial media industries must sort, categorize, and pass judgement on thousands of potential ideas for mass cultural products. Given this, much of the energy of cultural producers tends to focus on which ideas are worthy, which ones are unworthy, and why. In the case of the Hollywood , such judgements occur through a finite number of institutionalized languages for evalaution. In describing these processes, I propose that many organizations can be studied using what I call an “evaluation in interaction” approach. For organizations like those found in the Hollywood film industry, institutions for evaluating potential products and ideas become a central, organizing feature of the decision-making environment. Thus, the “evaluation in interaction” approach is concerned with discourses of evaluation as particular types of institutionalized discourse. Centrally, this approach holds that such discourses of evaluation are both historically formed and interactionally contingent. From this perspective, although conventions for evaluation are durable features of how judgements are made within organizations, the “firmness” of these conventions is softened by a couple of important realizations. The first realization is that even though institutions for evaluation are durable, they are neither natural nor inevitable. In the case of the Hollywood film industry, today’s dominant discourses of evaluation emerged out of particular historical and cultural moments during the 20 th century. Second, discourses of evaluation do not exist in isolation, but are reproduced within the context of particular 5 types of interactional institutions that predominate within institutions. I believe such an approach is useful because it helps integrate questions about institutionalization as a historical process with questions about institutions as coordinating features of contemporary organizations. Additionally, my exphasis on locating particular institutions of evaluation within particular institutions for interaction helps specify why particular kinds of institutionalized discourse rise to the surface in some contexts and not others. In the sections that follow, I first define what I refer to as “intuitive” and “technical” discourses of audience appeal and describe how Hollywood moviemakers use these modes of speaking to describe what kinds of products will work with audiences. I then discuss the theoretical foundations of the evaluation in interaction approach in more detail, exploring how it might address issues of evaluation in organizations. Intuitive and Technical Discourses in Hollywood In the course of my ethnographic research at different sites of Hollywood movie production, in observing how moviemakers talked at their professional conferences and in business school courses they taught, and in interviews with a variety of workers in the Hollywood film industry (studio executives, talent representatives, working screenwriters, script analysts, and audience researchers), I found two dominant, repetitive languages that moviemakers used in evaluating the viability of movie ideas. These two languages are founded upon two very different core principles of evaluation, or what Boltanski & Thévenot call “orders of worth” (1999). The first language, what I call an “intuitive” language of audience appeal, is based on a principle of “inspired judgment” (Boltanski & Thévenot 2006). According to the logic of this intuitive language, 6 individuals with good judgment can tell whether a movie idea is something that audiences want to see based on the perceived properties of a film idea itself. In this language, the locus of evaluating movie ideas lies with the individual and his or her “gut” sense of what works and what does not work. This is how one top studio executive I interviewed characterized his work, as a “gut game.” Moviemakers, however, do not continuously evaluate films in novel, idiosyncratic ways. Rather, they intuitively evaluate ideas using a set of terms that are meaningful in the world of Hollywood moviemaking: whether or not a film idea is “commercial,” whether or not it is “edgy,” or whether or not it is “artistic.” Although there is some variation in the precise meanings that these terms have for people working in the film industry, they nevertheless form common ideas that surround a film’s worth in the production cultures of Hollywood. This intuitive language for evaluating movies ultimately serves as a working knowledge about who will like particular types of films, how movies should be released and distributed, and how and to whom they should be marketed. A second language that moviemakers use in evaluating the appeal of films is what I call a “technical” language of evaluation. This language springs from the collected data and conclusions provided by audience research over several decades, and they establish a film’s worth according to measures that have the pretense of being “objective” and “factual”—removed from the biases of individuals or interests. Instruments and measures used to describe movie audiences are reified as capable means of getting at the real tastes of audiences and accurately evaluating the viability of movie ideas. This mode of evaluating movies is oriented around a typology (or cognitive map) of demographic groups that are viewed as predictive and meaningful. While spoken in a 7 language of dispassionate “objectivity,” this technical language nevertheless proceeds as a kind of Kuhnian “normal science” of audience tastes and spectatorship (Kuhn 1962), one where expected conclusions guide audience research and ultimately reaffirm the expected conclusions. These two languages each constitute shared, socially constructed forms of knowledge regarding why movies appeal to audiences. And, although these languages are used in evaluating specific movie ideas, they form ways of understanding why movies become successful in general and are thus very important in understanding how Hollywood (as a culture industry) works. As “discourses,” intuitive and technical languages of audience appeal are populated with terms and typologies that are meaningful within the social networks in which such languages are spoken (in this case, cultural producers in the Hollywood film industry). And, as shared grammars that order how the world seems to work for industry participants, these discourses also serve as powerful sources of constraint, limiting the potential ways in which movie ideas might be interpreted by moviemakers. Furthermore, these languages are “institutions” in that they have a durable, transposable quality, remaining meaningful across time and physical place (DiMaggio & Powell 1991). For Hollywood moviemakers, being fluent in these common modes of evaluating movie ideas becomes a prerequisite for membership in a network of fellow executives, producers, talent agents, writers, directors, and stars. Learning and practicing industry conventions for talking about what “works” with movie audiences is a skill that interns try to learn through osmosis, that screenwriters and directors use as they pitch and discuss movie ideas, and that executives take with them as they migrate from company to company. In 8 this way, these shared, taken-for-granted languages of “what works” form diffuse yet powerful discursive practices for those working in the film industry. How these institutionalized discourses of audience appeal take shape within the everyday, lived realities of work in the Hollywood film industry is the subject of this study. As Clifford Geertz put it, my aim here is to “thickly describe” how these discourses, as a set of shared meanings, help constitute the production cultures of contemporary Hollywood moviemaking (Geertz 1973). This study is partly an organizational study and partly a sociology of knowledge, although I see these two elements as closely intertwined. I argue that institutionalized discourses for talking about audience appeal (intuitive and technical) are intimately connected to the organizational contexts and interactional settings in which these discourses are used and ultimately reproduced. The linkages between the two can be thought of as flowing in both directions, from organizational context to discourses about audiences and from discourses about audiences to organizational context. In one sense, the varying organizational contexts of Hollywood (each dominated by their varying interactional games) all make possible and encourage particular kinds of institutionalized discourses about audience appeal. Organizations participating in the production of Hollywood movies can exist in different “organizational fields,” with each field involving different interactional rules for obtaining legitimacy in that field. The two ethnographic cases presented in Chapters 3 and 4, one dealing with a Hollywood talent firm and the other dealing with an audience research company, are examples of different organizations embedded in different organizational fields, each operating within its field’s rules for being perceived as legitimate among its peers. Amid these varying interactional rules for gaining and 9 maintaining legitimacy, particular kinds of discourse become more expedient than others. In this way, varying organizational contexts emphasize talking about audiences in particular ways and encourage different modes of evaluating motion picture ideas. Looking at this process in reverse, institutionalized discourses also help shape the interactional environment in which producers strive to obtain legitimacy among their peers. Discourses about audiences become functional in the multiple settings in which producers attempt to build their reputations. Here, moviemakers build legitimacy for themselves and their companies in part by drawing on legitimate ways of talking about audiences. Producers must therefore be fluent in the dominant ways of talking about audiences within their respective organizational field in order to be perceived as legitimate. Consequently, different ways of talking about audience appeal facilitate producers playing different kinds of “interactional games.” In the Goffmanian sense, different institutionalized discourses for talking about audiences encourage different kinds of “strategic self-presentation,” using different sets of rules (Goffman 1959). In the Hollywood talent industry, for example, the dominance of intuitive discourses for talking about audiences makes possible an environment in which legitimacy is highly dependent on presenting one’s interpersonal competence. This is because intuitive discourses of audience appeal, though they draw on shared terms with common meanings, ultimately prove very fluid in how they are deployed in everyday discussions of movie ideas. Technical discourses, on the other hand, prove much less fluid in how they can be interpreted. Measures and instruments as well as producers’ cognitive maps of demographic groups operate under the guise of being “objective facts” and are less malleable to situational circumstances. As a consequence, in the field of audience 10 research (where technical discourses about audiences are reproduced), researchers’ strategic interactional games are more focused upon reenacting preexisting instruments and measures. Despite self-doubts surrounding just how “objective” their research is, researchers attempt to strategically perform this objectivity in their presentation of research findings. Although intuitive and technical discourses of audience appeal are born out of different organizational circumstances and imply different principles for evaluating movie ideas, there is considerable overlap in what they do for the Hollywood film industry as a whole. Both languages of audience appeal are also languages of how audiences are different. The “truth” of these differences—creating causal links between “who moviegoers are” as distinct groups and then what sorts of movies these groups prefer is frequently experienced as a questionable endeavor by moviemakers themselves. Both intuitive and technical discourses about movie consumers (as shared forms of knowledge) are social constructs. Hollywood’s audience typologies (and the definition and relevance of particular groups within them) have changed over time. And, furthermore, both intuitive and technical knowledge about audiences are often challenged by confounding information that comes up in the course of evaluating movie ideas. A relevant question that I attempt to answer with this study, then, is how these dominant forms of knowledge are socially accomplished—how intuitive and technical ways of seeing audiences are reproduced in the everyday contexts of moviemakers’ work in Hollywood. 11 The Evaluation in Interaction Approach in Theoretical Context The “evaluation in interaction” approach that I use in this study illustrates how intuitive and technical discourses operate as historically formed and interactionally contingent cultural structures within the world of Hollywood moviemaking. Throughout this study, I refer to these languages as “institutionalized discourses”—a term that is used as shorthand for a specific type of cultural structure that exists within some contemporary organizations. The institutionalized discourses of evaluation that I examine here have several features, and draw theoretical inspiration from several conceptual frameworks. First, intuitive and technical institionalized discourses are durable cognitive structures. They have developed into instituions for talking about how and why motion picture ideas work. The durability of these discourses across the disperse locations and variegated players in the world of Hollywood filmmaking is best understood using insights from neoinstitutional theory. Second, these institutionalized discourses are, more specifically, modes of evaluation. They appeal to one of a finite number of core principles that can be used in making judgements about cultural objects (in this case, motion picture ideas). Given this, I borrow from Boltanski & Thévenot’s insights on how such discourses are organized around different “orders of worth.” Third, intuitive and technical discourses are epistemic in that they reference how moviemakers know what they know. These discourses are populated with a cosmology of terms (e.g. intuitive terms like “commercial” or “artistic” and technical terms like “youth” and “demographic”) that developed over the course of the 20 th century and now have particular meaning within the Hollywood film industry. These terms have emerged over time, forming the specific terminology used within Hollywood’s intuitive and technical modes of evaluation. The 12 development of intuitive and technical discourses (as two sets of terms) reveals to a history of competing epistemologies for understanding how and why movies appeal to audiences. Because these epistemologies of the audience can only be fully appreciated by looking at their historical genesis, I see Michel Foucault’s archaeological method as useful for understanding the emergence of these two sets of terms as competing “discursive formations.” Finally, both intutitive and technical discourses are used in interaction. They both prove to have different functions within the interactional contexts found in the contemporary Hollywood film industry. Here, I draw on recent sociological insights that see social interaction as comprising different cognitive institutions in its own right, shaping how other institutions (e.g., those having to do with collective discourses) might be used. The evaluation in interaction approach lies at the intersection of these conceptual frameworks. It attempts to dissect how institutionalized discourses are internally structured, as well as how they are deployed within specific interactional contexts found in particular types of organizational fields. The central features of these institutionalized discourses—their durability, the fact that they appeal to particular principles of evaluation, their epistemic quality, and the fact that they are deployed in interaction—are important facets of how these discourses emerge and are reproduced in the everyday worlds in which Hollywood moviemakers live and work. I explore how these different dimensions are important in understanding key processes surrounding intuitive and technical discourses, including their everyday reproduction and their historical genesis. 13 Instutitions as Durable Ways of Talking in Organizations With the coming of the postindustrial economy, organizational scholars have emphasized the importance of knowledge in the making and functioning of contemporary organizations. This rising “knowledge economy” in contemporary capitalism has come as corporations have become more dependent on global information flows, more sophisticated information technologies, and ever diversifying markets (Sklair 2001; Castells 2000). Here, managers and decision makers within the corporate world must assume the identity of “knowledge workers,” attempting to navigate the uncertainties of doing business under the mantras of “flexibility” and “creativity.” Key aspects of work for these corporate knowledge workers can involve formulating organizational strategies, understanding markets and consumer culture, and being familiar with specialized instruments and practices specific to a particular industry or company. Thus, important forms of knowledge in the contemporary corporation can be quite broad, and include strategies, managerial ethics, ways of interacting, as well as models of “how things work.” All of these are spawned within contemporary capitalist organizations as workers attempt to make decisions and manage uncertainty (Boltanski & Chiapello 2005). Much recent scholarship in the sociology of organizations looks at how these various kinds of knowledge become sedimented into practices that organize action. Organizational theorists think of these sedimented practices as “institutions,” defined as collectively shared, cognitive structures that take on “rulelike” status in social action, often across disperse settings (such as an entire industry) (DiMaggio & Powell 1991). These rules involve powerful, informal, tacit conventions for talk and action, and can come to dominate entire organizational fields. A common thread of recent work 14 influenced by this “new institutionalism” is that organizations can in fact be understood as constituted by shared institutionalized discourses and practices (Binder 2002; Armstrong 2002; Haveman & Rao 1997; Clemens 1993; Meyer & Rowan 1977). Institutions, which exist on a cognitive level (as a kind of shared culture), are fundamental to the functioning of organizations in a couple of ways. In one sense, institutions help organize the material world, as they help guide how resources are distributed in social life (Sewell 1992). Furthermore, institutions (as scripts for action) can become embodied in the material make-up of organizations, defining how physical artifacts and instruments are to be used (Swidler 2001). In this way, organizations are not objectively rational. Institutions are ways of culturally structuring action within organizations, and may or may not support the most optimal outcomes in terms of profitmaking. In contrast to neoclassical economics, the “new institutionalism” points to how dominant motivations for action are oriented around socially constructed notions of legitimacy rather than some singular pathway for profit-maximization (DiMaggio & Powell 1991). Institutionalized Discourses as Modes of Evaluation “Institutions” (as conceived by neoinstitional theorists) can involve a very diverse range of talk and practice that are used within organizations as sources of legitimation. Yet the coordination of many organizations (including those found in the culture industries) depends centrally on dicourses of evaluation. In the “evaluation in interaction” approach that I use here, I am specifically concerned with these evaluative forms of organizational coordination. 15 I draw on Boltanski & Thévenot’s (2006) ideas regarding “justification” as a basis for understanding these specific, evaluation-oriented discourses in organizations. In this approach, languages of evaluation are comprised of cognitive associations between the “things that count” when judgements are to be made, including objects, people, and concepts. In social life, the number of possible associative schemes is not infintite, but involve a finite number of modes for passing judgment. Much of the evaluation that occurs in human behavior is not idiosyncratic in style, but uses one of a set of languages that are “common” and thus “communicable” (2006: 32). Thus, the fact that Boltanski & Thévenot conceptualize evaluation as involving collective languages shows a point of commonality with neoinstitutional theory. Both perspectives, to some extent, examine the informal, durable, taken-for-granted scripts that structure collective discourse. But, for neoinstitutionalists, institutions are primarily defined by the fact that they have become sedimented into practice, and are socially meaningful only because they now operate as tacit rules for “how things are done.” Boltanski & Thévenot’s “justifications,” on the other hand, deal more specifically with collective evaluation. As such, they become socially meaningful not only because they present regularized modes of talking and passing judgement, but also because they can be used in argumentation. Given this, “justifications” cannot just involve any sort of collective language, but only languages that appeal to some sort of common principle along which evalaution proceeds. Boltanski & Thévenot think of these principles (e.g., “efficiency,” “profitability,” or “creativity”) as conventions for “establishing equivalence among beings,” and see them as governing one of six “orders of worth”—the common modes of evaluation in social life (2006: 140). 16 A second basic point of commonality between neoinstituionalism and Boltanski & Thévenot’s model goes to how collective langauges (whether discursive institutions in general or discourses of evaluation specifically) are used in fundamental ways to legitimize action. This proves true in the case of the Hollywood film industry, where moviemakers talk within either intuitive or technical lanaguages to to justify the extent to which movie ideas have “worth.” Although these two langauges appeal to different “common principles,” they both work as legitimate ways of conducting evaluation within this production culture. As Alain Desrosières (1998) puts it, The rationality of a decision, be it individual or collective, is linked to its ability to derive support from things that have a stable meaning, allowing comparisons to be made and eqivalences to be established. This requirement is valid for… persons seeking to construct, from common sense and objectivity, everything that enables us to guarantee a social existence transcending individual contingencies. A description can thus be assimilated into a story told by one person or by a group of people, a story sufficiently stable and objectified to be used again in different circumstances, and especially to support decisions, for oneself or others. (1998:6) Shared discourses provide a means for establishing equivalency and thus a means for making judgements appear “rational,” given that they end up appealing to collectively valorized common principle. Judgements thus are able to seem as thought they are outside the whims of individuals and within the boundaries of common criterion. However, in the context of organizations, neoinstitutional theory tends to see the connection between discursive institutions, the establishment of legitimacy, and action to be very forward-leaning. Under this conceptualization, discursive institutions have a tight correspondence with action. However, Thévenot & Boltanski’s work on justification suggests that institutionalized discourses having to do with evaluation are 17 much more shifty, contingent, and often prove to be much more decoupled from action than neoinstitutionalists would suggest. Here, legitimate forms of talk do not necessarily translate into corresponding forms of action. For Thévenot & Boltanski, there is inherent tension between the various “orders of worth” when applied to real-world situations. At times, this tension focuses on which order of worth applies to a given situation to begin with. Additionally, there are often “mitigating circumstances” that force compromises between principles of evaluation (2006: 150). This tendency for compromise leads to the proliferation of what Thévenot calls “composite organizations” (2001: 410). The members of these organizations “have to engage in different modes of coordination, depending on the configuration of the situation in which they find themselves… All organizations have to cope with critical tensions between different orders of worth” (ibid.). Composite organizations represent an adaptable, hybrid organizational form where organizations engage in evaluation drawing on two or more principals of evaluation simultaneously. Under the evaluation in interaction approach I propose here, Thévenot & Boltanski’s view of evaluative discourses is instructive in a couple of ways. First, it gives specificity to the kind of “institutionalized talk” that has to do with evalaution by showing how this talk is oriented around larger principles of evaluation. And, second, it moves beyond the “talk corresponds with action” paradigm that often typifies neoinstitutional insights. However, the evaluation in interaction approach helps refine Boltanski & Thévenot’s insights by exploring further how these discourses of evaluation are both epistemic and used in interaction. 18 Institutionalized Discourses as Organizational Epistemologies The discourses of evaluation described by Thévenot & Boltanski can be thought of in a couple of ways. In one sense, they provide a means for understanding the multiple scripts present in organizational coordination. But, in another sense, I would argue that they also exist as cognitive structures that answer the question of how organizations “know what they know.” Thus evaluative discourses can make up the epistemic foundations for decision-making in organizations and organizational fields—especially ones in which evaluation becomes a central organizational activity (like many of those in Hollywood). Different “orders of worth” represent different modes of assigning value to objects, but they also represent different systems of classification and fundamentally different ways of arriving at “truth.” Boltanski & Thévenot (2006) hint at the sometimes stark differences in epistemology implied by these various modes of evaluation. As they put it, …principles held in common are grounded in different common worlds. Something that belongs to and exists in one common world may be unknown in a different world. The world of inspiration, for example, is inhabited by demons and monsters, while the domestic world includes household pets, which go unrecognized in the civic world along with children and old people, and so on. Objects that constitute tools for making the worth of persons manifest in one world may not be taken into account at all in a different world.” (2006: 131). “Demons and monsters” may be meaningful terms for understanding “why things have value” under one order of worth, but irrelevant within another order of worth where an entirely different set of objects have meaning. For Thévenot & Boltanski, the various orders of worth do not come from nature, but instead create it (2006:136). In other words, orders of worth are ultimately kinds of 19 social constructs. But, this raises a series of questions surrounding “orders of worth” as shared cultural systems. For instance, why do particular terms belong (really) within particular orders of worth? In other words, why are “demons and monsters” necessarily some of the potential terms in the “inspired world,” but not others? How and why do some terms become especially relevant within a particular order (and within particular organizational fields operating within that order)? So, for example, when evalaution proceeds in the inspired world, why might “demons and monsters” be especially salient in some contexts but not others? In organizational analysis, such questions suggest the value of exploring the epistemological dimensions of orders of worth, focusing on how orders of worth (each with a constellation of specific terms) emerge as they do within particular organizational fields. If orders of worth do not naturally exist in nature, and particular orders of worth are not natural fits within particular organizational contexts, then an important task is to understand how and why particular orders of worth become relevant within particular organizational contexts. Answering this sort of question sheds light on how orders of worth (as epistemologies) come into existence. Thévenot sees this question as an important, if underexplored one, stating that “at the the level of societies, we are also occupied with another task: exploring the genesis of new legitimate modes of coordination and their compatibility or not with the grammar of orders of worth” (2001: 419). In organizational analysis, this task involves making linkages between the emergence of orders of worth on a societal level and following their manifestations within particular organizational contexts. The evaluation in interaction approach I advocate attempts to do this, tracing the emergence, coexistence, and/or 20 disappearance of orders of worth as modes of coordination within particular organizational fields (in this case, the Hollywood film industry). In carrying out this task, I see Foucault’s archaeological method as useful in understanding how orders of worth emerge as epistemological structures within organizations. This technique traces the development of terms, categories, and objects, looking at how they cluster into “discursive formations.” Such discursive formations are defined in their common movement, in “the way in which these different elements are related to one another” (Foucault 1972: 60). For Foucault, these discursive formations are indicative of underlying epistemologies, and, therefore, to see historical changes in the types of discursive formations used is to watch a shift in epistemic foundation. If we apply Foucault’s insights specifically to the discursive formations surrounding evaluation, there is a congruence between doing an “archeology of knowledge” (as Foucault understands it) and looking at the genesis of “orders of worth.” This synthesis—one I use in the evalaution in interaction approach presented here—advocates doing an archaeology of evalaution in organizational fields as a means of determining how an why some modes of evaluation have risen to the surface and not others. This approach has the potential of adding further specificity in understanding the anatomy of evaluation within organizations. Namely, tracing the emergence of “orders of worth” within organizational fields directs our attention to the specific terms and concepts that become meaningful within those modes of evaluation. By looking to this history, this allows an understanding of why some terms and objects (among many potential others) have become regularized terms of art within a particular organizational field. Thus, the evaluation in interaction approach is concerned with the genesis of 21 evaluation on a couple of levels: first, to see how general principles of evalaution emerged; and second, to see how these general modes of evalaution are specifically manifested through terms, concepts, and objects. This element of the evalaution in interaction approach helps reveal in more vivid detail how evaluative discourses in organizations are structured. Furthermore, it adds historical weight to the notion that particular modes of evaluation are not natural tendencies in organizations (i.e., not objectively rational), but have instead been socially constructed as legitimate modes of decision-making in particular organizational fields. Institutionalized Discourses in Interaction While discourses of evaluation are important structuring features of many aspects of social life, they do not exist in a vacuum. They must ultimately take form in interaction. As Boltanski & Thévenot put it: Since the principles of justice and the worlds in which they are realized are not attached to persons or groups but are instead embedded in situations, everyone encounters situations in daily life that arise from the various systems of justice, and in order to behave with naturalness, everyone has to be able to recognize these situations and adjust to them. People in whom this ability is lacking or impaired are deemed psychologically abnormal. (2006: 145) Thus, the widespread presence of evaluative discourses in social life demand the kind of interactional intelligence that Erving Goffman frequently identified. In the context of organizations, however, interaction frequently becomes more regularized and even institutionalized. Eliasoph & Lichterman (2003) see interaction as occuring among a finite number of institutionalized scripts for doing “groupness.” Here, interaction is organized as a set of cognitive schemas that people have for collectively 22 defining “what we are doing” as a group, what Eliasoph & Lichterman call “group style.” The finite number of scripts that people have for performing groups ultimately serve to place cultural constraints on collective action. In organizations, interactional institutions wark as focal points for constructing interpersonal legitimacy, and the interactional institutions that predominate within an organizational field confront individuals and firms who are trying to act in what they perceive to be their best interests (Zafirau 2008). Thus, organizational fields become strategic arenas in which people and firms attempt to legitmate themselves. Here, because institutions dictate ways of acting and talking that are constructed as legitimate within a particular organizational context, watching institutions happen involves watching people engage in what Erving Goffman called “strategic self-presentation” (see Goffman 1959). That is, people must ultimately perform institutions, and often do so for the purposes of appearing to be legitimate persons among their peers (who also recognize the shared significance that institutions have in a particular industry) (Zafirau 2008). Here, the “ideal self” that individuals seek to present to the world conforms with legitimate ways of talking and acting within organizations. This fits with a key assertion of Goffman’s dramaturgical model of social life—that people act strategically in attempting to conform to larger “situational properties”: Once you get the beast to desire socially delineated goals under the auspices of “self-interest,” you need only convince him to regulate his pursuits in accordance with an elaborate array of ground rules. (Important among these rules… are ‘situational properties,’ that is, standards of conduct through whose maintenance he exhibits regard for the current situation.) (Goffman 1967: 258-259) 23 Neoinstitutionalist theory amends Goffman’s insight by pointing to how these “situational properties” can and are shaped by wider symbolic systems for appearing legitimate within a particular industry. Thus, there are several levels at which “institutions” (broadly conceived) become important in organizations that engage in evaluation as a fundamental activity. One is at a “middle range” level, where they provide durable evaluative languages organized around orders of worth. These sorts of conventions are institutionalized discourses. But there are also interactional institutions that work at a more micro level, providing rules for how people legitimate themselves when interacting in their social networks. Eliasoph & Lichterman (2003) argue that institutions operating at these two levels can themselves interact with one another, and see the exploration of these processes as new horizon for research in the sociology of culture. And, various sociological analyses have begun to explore the broad question question of how discursive institutions (as shared cultural scripts and schemas) are used and interpreted within varying, micro-level interactional contexts (Hallett & Ventresca 2006; Magnuson 2005; Edelman 1992). Looking specifically at institutionalized discourses having to do with evaluation, Moody & Thévenot (2000) look at how justifications operating within various orders of worth can be used strategically. However, the links between institutionalized discourses of evaluation, organizational fields, and strategic interaction within organizations need to be fleshed out more precisely. In this project, I attempt to view Hollywood moviemaking as an evaluative process at the intersection of these different kinds of institutions (the discursive kind and the interactional kind). Here, I conceptualize moviemakers as not only producing culture 24 for audiences, but also engaging in this production amid multiple cultural constraints. These constraints include the overlapping institutionalized discourses they have for evaluating movie ideas using langauges of audience appeal, which are both intuitive and technical. And, they simultaneously include the interactional institutions that moviemakers strive to adhere to as they strategically present themselves to business associates. Overall, in detailing the evalaution in interaction approach, I am attempting (in a specific sense) to better theorize processes of popular culture production via the mass media as well as (in a broader sense) begin to offer a better theorization of organizations focused on evaluation. I see such organizations as complex, using multiple and intersecting modes of coordination, and operating within varied cultural structures with varied forms of constraint. The evaluation in interaction approach offers a deeper conceptualization of such organizations in couple of ways. First, by seeking to trace the historical conditions under which discourses of evaluation become salient within particular organizational fields, it highlights how such organizations are “rational” only in the sense that they utilize legitimate modes of evaluation that were socially constructed over time. Along these lines, it also helps answer why particular terms, concepts, and objects became meaningful within particular organizational fields, offering a further level of specificity. And, second, it further refines Boltanski & Thévenot’s model of evaluation for organizational analysis. The “evalaution in interaction” approach’s attention to the constraining effects of interactional institutions within organizations places it between neoinstitutionalism and Boltanski & Thévenot’s ideas about justification. A primary 25 assertion made by Boltanski & Thévenot regarding organizational analysis is that “we need a notion of coordination which is more open to uncertainty, critical tensions and creative arrangements than the ideas of stabilized and reproductive orders” (2001: 406). Accordingly, organizations that deal with evaluation are much less regularized than neoinstituionalists would suggest. However, I argue that this view of the organization as “fluid” needs to be tempered somewhat. Although discourses of evalaution by themselves tend to provide a fluid and ultimately indefinite means of coordination, they also operate within institutionalized forms of interaction which can provide a great deal of certainty, helping to determine when and how evaluative discourses are deployed. Here, I see Boltanski & Thévenot’s emphasis on “compromise” between orders of worth as overstated (at least in the case of Hollywood). When interactional institutions within organizational fields help determine whether and which orders of worth are used, particular organizational conditions will tend to correspond with particular “orders of worth.” Thus, institutionalized rules of interaction may in fact reproduce dominant modes of evaluation within a particular organizational context. Having theorized the evaluation in interaction approach on a general level, I now turn to the specific question of popular culture production, focusing on what this approach adds to studies in this area. Making Popular Culture and Audiences Gil Cohen, the former studio head who taught a class I observed on the movie business, once made the following observation: “Academics are very good at describing the structure of the industry. But what they don’t do very well is tell you how it actually 26 works.” Over the course of my fieldwork, I gradually began to understand what Gil Cohen had meant. The Hollywood game was filled with many unspoken rules for how business was to be done that were not apparent simply by looking at how revenue flowed into the studios, the industry’s oligopolistic market structure, or the various technological developments thought to be changing how movies will be distributed and watched. While all of these things mattered, there was also a powerful subculture that ran through the business of making movies—one that at times failed to make sense to participants, but knowledge of which was required for true “insider status.” I see Cohen’s statement as having both theoretical and methodological implications for research on “the culture industries.” A great deal of research on how the machinery of popular culture production operates proceeds from a strong “structuralist” bent. Some analyses of the media industries, for example, tend to reduce popular culture production to a unitary drive toward profit-seeking on the part of producers. And indeed, the desire of film studios to increase their profits is certainly the motive articulated by moviemakers themselves. But as Michel Foucault suggests, discourses—whether they be discourses surrounding sexuality, criminality, or motion picture audiences—are not reducible to economic interests (see Foucault 1980, 1978, 1977). Instead, the motivations that people attribute to themselves are also mediated through larger discourses. While the search for new revenues and the lure of finding new “windows of exploitation” are a basis for relying on particular explanations of movie audiences, the way that knowledge about audiences has and is developing represents but one possible version of Hollywood’s search for new profits. What if, for example, instead spending a great deal of money on developing “mobile delivery platforms” (iPods and cell phones), studios 27 devoted resources toward enhancing the theater experience? Or, if instead of greenlighting expensive remakes of past blockbuster successes that reach their apex on the opening weekend, they focused on making a larger number of less expensive films involving original material that attracted audiences over time? These different visions of Hollywood’s “business model” would depend on different institutionalized discourses of audience appeal. Given this, uncomplicated assertions by those in the business of Hollywood that “people are just getting what they want” are questionable from the vantage point of this study. It is not as though those “wants” are experienced as immediately transparent to those working in the film industry. Moviemakers have, at best, only imperfect lines of communication with their audiences, and considerable work is done in attempting the uncertain task of understanding Hollywood’s consumers. Thus, “what the audience wants” is a matter of constructed perception. Instead of making the claim that “people just get what they want,” it would be more precise to say that people are getting one possible version of what those in the movie business talk about as being profitable. Intuitive and technical languages of audience appeal, as dominant ways that moviemakers talk about movie ideas, become the Hollywood movie industry’s working theories of why movies will be profitable and twin modes of determining whether ideas have worth. But although these languages may be spoken as confident assertions about why movie audiences work the way that they do, examples abound of how these languages fail to account for the industry’s many misapproximations. In one vein, moviemakers’ shared intuition has sometimes resulted in sub-par—even disasterous— box office returns. Regular assumptions embedded in this language are brought into 28 question in these instances. Movies thought to have strong “commercial” appeal often become lame box office flops, while movies thought to lack widespread “commercial” appeal become surprise hits. Or, elements commonly thought to predispose a film to commercial success (e.g., star power) regularly fail to guarantee success in the final analysis (De Vany & Walls 2002). On the other hand, Hollywood’s technical languages for understanding audience appeal, geared toward scientifically objectifying audience tastes, often notably fail as well. Breaking a movie down into its various plot, visual, and narrative elements and then systematically measuring how these elements perform with test audiences is a strategy that has also yielded inconsistent results at the box office. All of this illustrates how the Hollywood film industry’s strategies for understanding and talking about its audiences often fail to produce the optimal financial outcomes. And, as this study highlights, moviemakers themselves are often the first to admit that their industry’s strategies are not optimal, much less driven by rational decision-making. Rather than attempting to explain popular culture production primarily in terms of the essential economic motives of media producers, I believe that a more fruitful approach is to emphasize the ideational and discursive aspects of media production. From this perspective, media audiences do not exist so much as real entities “out there” in the world for producers to discover and understand. Instead, audiences exist in the shared imagination of media producers. Although I argue that the ways in which audiences are now spoken about have become institutionalized in contemporary Hollywood, these ways of seeing audiences have only come into existence relatively recently in historical time (an evolution which I partially trace in Chapter 2). That people would come to be seen as audiences in the contemporary sense at all (i.e., as distinct groups with authentic 29 characteristics that can be appealed to) presupposed a shift in how human cultures were viewed. According to Jon Cruz, this involved a very modern way of viewing culture that was born in the 19 th century, one in which a new sort of gaze from interpreter to interpreted came into existence. For Cruz, the birth of this modernist “cultural interpretation” involved the work of proto-anthropologists as they began looking into the cultures of various “others” for those “cultural performances, practices, goods, and objects that could be studied in order to uncover a posited deeper reality” of a particular group (Cruz 1999: 7). This way of seeing groups of people as real, discrete entities with real, decipherable characteristics laid the groundwork for cultural producers in the entertainment industries to also eventually see audiences as “real.” It made possible some of the key assumptions currently held by moviemakers: that audience groups actually exist and their tastes can in fact be meaningfully interpreted. Given these assumptions, it came to make sense that moviemakers interested in making financially successful movies might strive for deeper knowledge of the groups of people who consumed their products. For much of the early history of electronic entertainment media in the United States, this kind of interpretation proceeded based solely on the intuitive guesswork of producers themselves. But other aspects of modern life would eventually further elaborate on the ways in which media producers would interpret audiences. One of these forces involved an increasing quantification of social life. This corresponded with the rise of the modern state and modern social science (see Foucault 1978), both of which collaborated in measuring populations by counting them. From quantification were born notions of “normalcy” and “typicality,” ideas that could add further specificity to distinct 30 groups (such as they were). When moviemakers began to utilize social scientific methods to interpret moviegoers (beginning in the 1930’s and intensifying through the present day), audiences would not only be seen as real, distinct groups but as real, distinct groups that also had “typical” behaviors. This additional wrinkle in how audiences could be interpreted by moviemakers corresponded with a rapid influx of marketing professionals into the entertainment business from the 1970’s forward (see Napoli 2003; Acland 2003; Wyatt 1994). These new professionals in the entertainment industry helped institutionalize technical ways of talking about audiences. Finally, the rise of large-scale media companies in the early 20 th century added further impetus for moviemakers to use modern modes of cultural interpretation in their attempts to understand audiences. Movie production was initially a largely organic activity of relatively poor, largely immigrant urban dwellers (Abel 2006; Ross 1998). Short films were made on a local basis, and shown to local audiences who likely lived close to where movies were made. As venture capitalists saw the potential of the movies and tried refashion them as a mass leisure activity, the forces of conglomeration set in and eventually resulted in the concentration of movie production among a relatively few large studios by the 1930’s (Balio 1993). Motion picture studios were unabashedly for- profit enterprises, more bureaucratic, and attempted to make movies for much larger swaths of people than did the earliest moviemaking enterprises. These large, corporate entities were more removed from the great “mass audience” to whom they now sought to appeal. With the creation of this divide between the world of production and the worlds in which audiences lived, modern modes of attempting to understand audiences and what appealed to them gained favor. 31 Media Production and the Idea of the Audience Popular culture production today powerfully involves notions of “the audience,” ideas that at times consume moviemakers’ work while at other times enter into the production process more implicitly. A number of ethnographic and interview studies of entertainment media production have shown how producers go about their work with assumptions of who their audiences are and what sort of content they desire (Mayer 2008a; Grindstaff 2002; Gamson 1998; Ettema et al. 1994; Ahlkvist & Faulkner 2002; Gitlin 1983; Cantor & Jones 1983; Espinosa 1982; Cantor 1971). Sometimes, these assumptions refer to particular demographic groups, such that producers have a specific demographic in mind as they go about making content. At other times, assumptions about audiences involve more universal appraisals of “what people want” that cut across demographic lines. Here, producers might draw upon notions of “human nature” in justifying their assumptions about audience tastes. Whatever assumptions producers have about audiences, attempting to obtain usable knowledge about media consumers can be experienced as a conscious part of producing entertainment content. That is, producers may experience this as work. In the cases of many Hollywood moviemakers, the labor required to understand audiences is made all the more challenging because of the large social distance that producers see as existing between themselves and moviegoers at large. Places involved in the greater Hollywood production milieu include office towers in Century City and New York, studio lots, trendy Los Angeles area restaurants, and homes in Beverly Hills, Santa Monica, or Malibu. Furthermore, people working in Hollywood might spend much of their workaholic existences cloistered among others who spend most of their time in 32 these exact same places (Zafirau 2009). Media scholars drawing on the work of Pierre Bourdieu see this social cloisteredness as reinforcing a kind of “habitus,” where producers (whether they be moviemakers, television producers, or journalists) are removed from their audiences by substantial gaps in economic and cultural capital (see Benson & Neveu 2005; Bourdieu 1993, 1977). It is within this relatively exclusive social network, with its undertone of collective anxiety about whether or not movies will “work,” that producers use institutionalized discourses of audience appeal. It is not only that these institutionalized discourses are used in production activities. Equally important is how in using technical and intuitive modes of talking about audiences, moviemakers reproduce their audiences as social constructions. Ideas about movie audiences become real to moviemakers in the way that they are collectively imagined within “the industry,” and in this way audiences also exist as a kind of shared, binding knowledge that ties together the disperse world movie production. To gain further insight into this world, I return briefly to the “structural story” of the Hollywood film industry, with an eye toward describing how collective knowledge about audiences circulates among key entities involved in the process of making movies. Hollywood Moviemaking: Major Players and Knowledge Flows Both scholars and non-academic inside observers of Hollywood note the complexity of the contemporary movie production process (Wasko 2003). In Hollywood’s “classic” era (c. 1930-1950), this process was considerably less complex. During that time, studios encompassed all elements of the filmmaking process under one corporate roof. Writers, actors, and directors worked for only one studio and were tightly controlled by studio 33 heads (Gamson 1994). In addition, studios each owned their own chain of exhibitors, which provided a venue through which studios could more rigidly control distribution (Balio 1993). Today’s Hollywood is (organizationally speaking) less centralized, and the machinery of contemporary film production has many moving parts. Significantly, “talent” is no longer under the direct control of the studios, but is bought and sold through talent agencies and management companies. Furthermore, following a U.S. Supreme Court ruling in 1948, exhibitors were forced to operate independently of studios. Figure 1 provides an overview of where and between whom knowledge flows in the Hollywood film production process. This diagram is meant only to provide a rough (if oversimplified) sketch of how this process works, and of how the film production process maps onto the circulation of knowledge about audiences. Although this diagram has been reconstructed using scholarly sources and my own observations, the picture drawn here is not representative how this world always functions. The general trajectory of the film production process is as follows. First, ideas are formulated and reworked by screenwriters and their agents or managers. My fieldwork at Limelight (a talent management firm) confirmed what various agents, managers, and writers also stated at conventions and in seminars: that the conceptualization and revision of scripts involves extensive collaboration. After an idea has undergone its initial conceptualization (as embodied in a finished script), writers, agents, managers, and lawyers attempt to move this idea, as formulated, through the studio system. Often, they do this by attempting to sell (or “spec”) a script to a studio or production company. The process by which scripts are acquired is a highly selective one, and it is estimated that 34 only approximately 5% of all scripts “speced” to studios receive more than a second look (Taylor 1999). If a script is acquired by a studio, a series of decisions are made within the studio that usually alter the original idea—sometimes considerably. The individuals involved here include those working in the studio’s story department, various creative executives, and other development personnel. At the end of this process (or “development”), comes a final decision about whether a movie should made or not, which involves members of a prospective film’s “greenlight committee.” This committee typically includes an array of executives from the studio that will finance the film. Ideally for the studio, the entire cast, crew and production schedule for the film are now set. Finally, important decisions are made as to who the target(s) are for a particular film, and how these target(s) might be convinced that the movie is worth seeing. These sorts of decisions are likely made early on in the process—or at least well before the film is actually completed—and are documented in the studio’s written marketing plan for a particular film. As a film finally nears completion, however, the studio’s marketing wing moves into full action and becomes pivotal. It advises on when a film should be released, utilizes audience research to weigh the advantages and disadvantages of pursuing different target audiences, and designs “creative advertising” and publicity campaigns to attempt to bring as many viewers into theaters as possible. Although each of Hollywood’s many organizations may have very different businesses, they are connected in a close, self-scrutinizing world. The presence of dense interpersonal networks—and their importance for the operation of film business—are underscored by the close geographic proximity that major studios, production companies, 35 and agencies all share (see Scott 2005, 2004). “Hollywood,” which is frequently used here (and elsewhere) as a shorthand term for “the U.S. film and television industry,” is actually no longer centered within the physical location of Hollywood. Beginning several decades ago, the film industry’s major companies began moving to surrounding areas of Los Angeles. Studio lots are found in Burbank and Century City, and much of the industry’s business is now conducted in Beverly Hills, and, specifically, along the Wilshire Blvd. corridor. This area houses the industry’s five major talent agencies, the offices of various production companies, and the expensive restaurants where business meetings are often held. Figure 1: The circulation of audience knowledge in the Hollywood system: Major players and processes 36 Hollywood’s dense interpersonal networks are a backdrop for the industry’s continuous internal discussions of itself. Major trade publications, including Variety (whose main offices are also located on Wilshire Blvd.), reflect much of this collective discussion in their daily issues, which are disseminated widely to those working in Hollywood. As shown in Figure 1, Hollywood’s various trade publications are important media through which knowledge about audiences is constructed, reinscribed in discourse, and widely disseminated to an organizationally decentralized film industry. As Acland suggests in his study of the discourses about film exhibition that run through the film industry, Hollywood’s trade publications have become important as “vehicles of debate, analysis, and information” and provide a “point of intersection for an increasingly dispersed industrial sector” (2003: 21). My ethnographic observations have provided firsthand evidence that this is the case. At Limelight Entertainment, the talent firm where I worked and observed for several months, all of the firm’s talent managers received daily copies of Variety and most received the Hollywood Reporter. For many of these managers, one’s day begins by reading the “trades,” and these publications are continually mined for information about deals, potential clients, reporting about current clients, and articles dealing with the industry more broadly. Thus, although many different entities participate in making movies, they share a common forum for discussing audiences by way of the trades. The Chapters to Come The remaining chapters of this study delve into how intuitive and technical modes of talking about audiences emerged as they have in the contemporary motion picture 37 industry, and how these discourses of evaluation are reproduced as shared knowledge in the everyday, lived worlds of Hollywood moviemaking. From the fall of 2005 to early 2007, I gathered ethnographic, interview, and questionnaire data from a number of sources. This endeavor involved months of participant observation research at “Limelight Entertainment” (a major talent management company) and “Strategic MediaVision” (an audience research firm servicing a variety of media companies, including motion picture studios); participant observation in courses on the movie business offered through a major business school in the Los Angeles area, all of which were taught by figures who were working or had worked at the upper echelons of Hollywood; interviews with approximately 30 motion picture industry professionals, including studio executives, talent agents and managers, audience researchers, working screenwriters, and studio script consultants; attendance at screenwriting and other entertainment industry conferences; participant observation at a number of audience test screenings; and the gathering of questionnaire responses from working studio executives. Over the course of this research process, I gradually developed insights into how shared languages of audiences existed across disparate individuals and locations in Hollywood (see the Methodological Appendix for more details on the research process). As a way presenting Hollywood’s collective audience discourses in historical view, Chapter 2 partially traces the development of both intuitive and technical discourses about audiences as modes of evaluation within Hollywood. I focus on key periods of the film industry’s history, drawing on a large sample of issues of Variety (Hollywood’s leading trade publication) from years 1950 to 1955 and from 1970 to 2005 in order to trace the evolution of audience discourses over the last several decades. These 38 periods represent important ones in Hollywood’s changing ways of talking about audiences, helped usher in the adoption and eventual full-scale institutionalization of market research into moviemaking. These shifting discourses of evaluation within the film industry also occurred against the backdrop of larger societal changes pushed changes within Hollywood’s internal decision-making environment. I argue that as technical modes of evaluation began to appear, so did shifts in Hollywood’s overall logic of audiences asr previously dominant intuitive understandings of audiences were revised. Chapters 3 and 4 explore how intuitive and technical discourses (respectively) are reproduced—and made possible—against the backdrop of particular types of organizational contexts found in contemporary Hollywood (see Table 1). These organizational contexts are defined by the structures of the organizational fields in which these different companies operate, which in turn spawn dominant conventions for how producers within these fields seek to legitimate themselves on an interactional level. The production of different institutionalized discourses of evaluation becomes possible amid these varying interactional conditions. ______________________________________________________________________________________ Table 1: Institutionalized audience discourses and the organizational contexts in which they are reproduced ______________________________________________________________________________________ Site Organizational field Locus of interactional legitimacy Institutionalized within organization discourse about audience Limelight hierarchical interpersonal “reputation work” intuitive Strategic flatter competence with measures/ instruments technical MediaVision ________________________________________________________________________ 39 In Chapter 3, I unpack key terms and concepts within intuitive modes of evalaution. Here, I look at how movie ideas are perceived as “commercial,” “edgy,” and “artistic,” and suggest how these distinctions necessarily limit how movie ideas can be understood for those working in the motion picture industry. I then show how these sorts of discourses are reproduced amid the everyday interactional conditions found at Limelight Entertainment. Here, talent managers and producers strive to present themselves as “power players” within a talent representation business that is very hierarchical—separated between “insiders” and “all the rest.” Within these ongoing interactional games, or an interpersonal kind of “reputation work,” intuitive discourses of evalaution become more expedient for advancing talent managers’ interpersonal agendas. In attempting to maintain the perception that they are “power players” in Hollywood’s business networks (i.e., that they maintain viable contacts with A-list stars, top directors, and movie executives), intuitive ways of talking about “what is going to work with audiences” can be adapted to fit the interactional exigencies of talent managers. In a parallel analysis, Chapter 4 looks critically at the collected assertions of “scientific” audience research. I show how this knowledge relies on what I call “demographic types,” which have the effect of routinizing how moviemakers see audiences as distinct groups with their own characteristics and tastes. Here, motion picture audience research bears commonality with other kinds of scientific polling in that it allows for notions of “normality” and “typicality” to surround particular populations (see Igo 2007). Here, looking at the case of Strategic MediaVision, I highlight the dominance of measures and technical instruments in organizing work for audience researchers. As a consequence, or because these measures are taken to be objective and 40 rational, the field of audience research firms is less hierarchical than the organizational field found in the talent industry. Building reputation is still just as important for audience researchers as it is for talent representatives, but the type of interactional game is profoundly different. Here, reputations proceed not so much from whether or not people are perceived as “connected” within hierarchical social networks, but based upon their competence with standardized methods of audience measurement used in the entertainment industries. In this context, audience researchers seek to perform objectivity by reproducing dominant technical discourses about audiences, even when researchers see this “objectivity” as an impossible goal. Familiar technical stories about audiences can thus provide audience researchers with convenient, face-saving answers when no usuable answers emerge from research itself. Finally, Chapter 5 further considers the implications of this study for organizational research. Here, I discuss further how the evalaution in interaction approach might be applied to other cases, focusing on how evalaution (as a means of organizational coordination) might be embedded in other kinds of interactional contexts, found in varying organizational fields. 41 Chapter 2: The Evolution of Audience Knowledge in Hollywood In the years following World War II, Hollywood operated under a different business model than it does today. Movies were made and distributed to a larger number of theaters. Strategies for releasing movies relied on keeping them in theaters for a long duration, giving them time to marinate and build in popularity. The conventional wisdom on what the calendar for box office receipts would look like held that the peak season for returns was the Christmas and New Years holiday season. Dominant understandings of the motion picture audience still tended to equate American moviegoers with the public at large, only loosely differentiated by demographic characteristics that were not yet fully crystallized for moviemakers. Today, Hollywood’s business model has fundamentally changed. The film industry remains closely wedded to the contemporary blockbuster, films that are tightly focused on their big opening weekends and that taper off very quickly. The heaviest seasonal target for movie studios is the summer, when schools let out. And, finally, the motion picture industry now recognizes an audience beyond “the public at large.” There are a variety of precisely defined segments that are meaningful in the parlance of Hollywood moviemakers. Within this cognitive map of the American movie audience, there remains a notion that some films have “broad” appeal. Yet, these films are seen as largely appealing to people under age 25, and playing disproportionately well with less educated audiences. And, Hollywood also recognizes a variety of “niche” audiences. Depending on the niche, such audiences can either compliment “mass audiences” or be 42 defined by virtue of their opposition to “mass audiences.” Important segments in this regard include the “independent” film audience and a variety of “ethnic” audiences This paradigm shift in the way Hollywood thought about its audiences occurred against the backdrop of wider changes in American society and culture over the late twentieth century. These cultural shifts bled into internal debates occuring within the film industry as it attempted to adapt to the changes that it perceived as happening among the American movie audience. First among these was the birth of television in the late 1940’s, and the changing patterns of leisuire time that seemed to correspond with its diffusion. In 1946, there were an estimated 8,000 television sets in use in the United States, a figure that increased to approximately 3.9 million sets by 1950, and 10 milion by 1951 (quoted from Baumann 2007). This rapid diffusion of the television and the increasing amount of content that was available through this medium meant (in theory) that people might spend more of their spare leisure time at home and less of it engaging in other activities, including going to the movies. Although the trends of rising television use and declining theater attendance did not exist in some sort of neat, one-to-one inverse relationship with each other, the long-term pattern resulted in a reduced percentage of potential moviegoers actually going to movie theaters. Second, important demographic changes occurred in the post-War years. The years between 1946 and 1964 saw a spike in the American fertility rate to above 100 births per 1,000 women, the only time since that this rate has been so high. This was the post-War Baby Boom, which saw the births of approximately 80 million new children (U.S. National Center for Health Statistics 2001). As the proportion of the population 43 under the age of 18 expanded, so did the proportion of potential moviegoers get relatively younger. Accompanying this demographic trend toward youth was the changing cultural meaning of “youth” over the mid-twentieth century. In the United States, social and cultural upheavals tied to this young people resulted in their shifting symbolic status in American society. Increasingly, young people were tied to what Stanley Cohen called “moral panics,” periods in which a “person or group of persons emerges to become defined as a threat to societal values and interests” (1972: 1). A whole line of sociological work—from Cohen’s study of Great Britain’s mods and rockers (1972), to Hebdige’s (1979) explanation of 1970’s punk subculture, to Gaines’ (1990) description of heavy metal kids during the 1980’s, to Sternheimer’s (2006) examination of popular myths surrounding today’s teens—converges on how youth subcultures in the late twentieth century frequently became flashpoints for cultural anxieties. Non-conformist modes of dress, musical styles, and other dangers perceived to be lurking within the lifestyles of young people helped crystallize a sense of danger in successive panics. In his analysis of the hippie counter-culture of the 1960’s and 1970’s, for example, Jock Young (1970) saw that “the argot, way of dress, taste in music, interest in the occult, pattern of drug use, attitutdes to conventional employment,” and “philosophy of life” all made hippies “a ready target for moral indignation” within wider society (148-149). In defining such dangers and connecting them to young people, “youth” became more visible and meaningful as a category of people within society at large and within Hollywood specifically. This category helped Hollywood differentiate its audiences as 44 technical discourses of evaluation gained prominence alongside older intuitive discourses of evaluation. Finally, the emergence of technical discourses of evaluation in the film industry during the 1930’s and their rapid intensification after World War II was closely correlated with the rise of consumer marketing within American capitalism, which in turn had its roots in statistically-driven social science. Centrally, statistical techniques relied on the idea of an “average.” In his history of statistics, Desrosières (1998) shows how this notion has involved an ongoing debate going back to medieval Europe over the “reality of groups.” This debate generally has not involved the reality of the average as a statistical technique, but rather the extent to which it is used to represent a meaningful aggregation of attributes among human groups. In light of this history, Desrosières ultimately concludes that “aggregates do exist” as collective constructs, but only when such aggregates become useful objects in the social world (1998: 101). Over the course of the twentieth century, statistical techniques helped Hollywood shape a number of its now meaningful audience categories. “Youth,” social classes, women and men, for example, would come to be seen much more as a meaningful wholes with aggregate tendencies over this time period. These categories provided a topography of terms that could be used within an ascendant technical mode of evaluation, giving Hollywood moviemakers a specific language for deciding whether and how movie ideas might have worth. How movies have changed since the end of World War II has been the subject of considerable attention from both scholars and observers from within Hollywood itself. Often, these changes in how movies are sold to audiences are attributed to natural 45 evolutions in technology, broader cultural changes, the resulting shifts in movie audiences, and moviemakers’ natural recognition of these changes in appealing to their markets. What is less emphasized is the power of how movie audiences are spoken about within the film industry itself. Rather than taking the actions of moviemakers as a “dependent variable,” I suggest that the collective ideas of moviemakers themselves shape the ways that the movie industry continues to change. This is not to say that audiences do not have real characteristics, or that audiences do not really change, or that larger social and technological trends do not encourage/ discourage changes in the constitution of audiences. Rather, it is to point out that “real” phenomena surrounding audiences are first mediated through media producers, who must perceive what is happening in their markets using available techniques for making and interpreting audience knowledge. What I do argue here is that this knowledge is at least partly independent of “real” trends. As I discuss in Chapters 3 and 4, dominant ways of understanding audiences are made possible within particular organizational contexts that can prove decisive in maintaining these models of how audiences work, even in the face of doubts and contrary evidence. Following Intuitive and Technical Knowledge of Audience Appeal Over Time In the rest of this chapter, I describe how Hollywood moviemakers’ discourses of audience appeal have developed since the 1950’s. The transition from Hollywood’s model of movie audiences right after World War II to its contemporary model involved a series of important discursive shifts. Tracing how discourses of movie audiences developed over this time span involves what Michel Foucault saw as an “archaeological” 46 project (Foucault 1972). That is, by showing how audiences were spoken about at different points in time, including how audience categories originated and died, how they multiplied and folded into each other, we can see the genesis of current ways of talking about audiences within the Hollywood film industry. In this way, it is possible to see audience knowledge not simply as an objective reflection of contemporary realities, but as a set of ideas that have, at times, been contested. Additionally, such an archaeological testifies to the power of audience discourses in organizing Hollywood movie production because collective audience discourses dictate the parameters of what producers are capable of saying about audiences. How producers were capable of talking about audiences over the latter half of the twentieth century shows not only how moviemakers talked about audiences, but also how they helped give life to various identities of movie consumers. To undertake this history of audience discourses in the Hollywood film industry, I performed a content analysis of all articles published in relevant sections of Variety during the months of May, June and July from 1950 to 1955, and from 1970 to 2005. 2 Variety is the film industry’s leading trade publication and a central location through which common audience knowledge is reported, constructed, and disseminated to an organizationally disperse film industry. As Acland suggests in his study of discourses about film exhibition that run through the film industry, Hollywood’s trade publications have become important as “vehicles of debate, analysis, and information” and provide a “point of intersection for an increasingly dispersed industrial sector” (Acland 2003). My 2 Relevant sections of Variety included the front section (which often contained news specific to the film industry) and the film section. Over the years, Variety has also published news specific to other parts of the entertainment industries, including radio, home-video, television, and legit (i.e., theater). Refer to the Methodological Appendix for more on how this was done. 47 ethnographic observations at various productions sites and in business courses on the film industry all confirmed that “the trades” were consistently read, especially among talent agents and managers (but also among marketers as well). For some, reading these publications (including not only Variety, but the Hollywood Reporter along with several online sources) formed a routine part of the workday. This was seen as vital for moviemakers, as they used the trades to keep abreast of deals between other parties in the industry, movements of their colleagues from company to company, and overall trends occurring within the industry. Admittedly, this sort of analysis is necessarily limited in what it can reveal about “the industry” as a whole. Newspapers only report what their reporters and editors deem relevant to report, thus making it problematic to say that any news publication is a complete representation of events and trends occurring in the world. Furthermore, very little can be said about how readers of newspapers interpret these events and trends without actually talking to readers. Nevertheless, newspapers do offer a common frame of reference for readers. In the entertainment industries, where trade publications like Variety are dutifully read as a matter of occupational necessity, they supply what is perhaps the only common frame of reference. This was an oft-cited reason moviemakers gave for reading the trades when the subject came up in the course of my fieldwork: “to know what people are talking about.” That is, trade publications had relevance precisely because they allowed readers to be in “the know” with their peers, to be able to be conversant in what was happening in the industry as a whole (as reported through trade publications). 48 Changing Ways of Understanding Audiences, Changing Discourses of Audience Appeal Variety’s reporting on the Hollywood film industry from 1950 to 2005 reveals (often implicitly) significant changes in how moviemakers attempted to make sense of audience tastes over the last half of the twentieth century. Ever since moviemaking emerged as a profit-driven enterprise, making movies has involved a great deal of intuitive guesswork as to what audiences would find appealing. Studio executives have long justified their positions within the industry by citing their own gifts of perception in choosing movie hits. But, beginning in the 1930’s, technical ways of understanding audiences were introduced as an additional, sometimes rival, way of understanding why movies would work with audiences. My analysis below points to how discourse steeped in these new techniques intensified during the 1970’s (especially) and became thoroughly integrated into how audiences were understood by the 1980’s. This shift corresponded with new technologies of audience measurement, the influence of consumer marketing from outside the film industry, and the rise of specialized organizations dealing with audience research (like the one described in Chapter 4). Instead of displacing the intuitive prerogatives of moviemakers altogether, technical discourses of audience appeal simply added further complexity—both shaping and being shaped by intuitive discourses of how audiences work. What has resulted is a hybrid form of knowledge about motion picture audiences that now circulates within the film industry, where intuitive guesswork and technical discourses have interacted to produce a working model of audience appeal that links types of movies to types of media consumers. 49 Overall, this map of types of movies matching types of consumers can be described in terms of a shift in moviemakers’ perceptions regarding who constituted the American moviegoing public. This shift involved a general movement from seeing a less differentiated “mass audience” comprised of adults and “families” to moviemakers eventually seeing audiences as both younger and more differentiated. Not only did the center of gravity move away from adults (broadly understood) and toward moviegoers under age 25, but it also came to see various niches that were describable as aged, classed, and raced demographic segments. These groups were formerly folded into the “mass,” but were now given identities that eventually became meaningful audience categories for moviemakers. As these groups became talked about with greater intensity, they set up several fundamental oppositions that dominate moviemakers’ understandings of which movies appeal to which moviegoers, including “young” versus “old” and “blockbuster” versus “independent.” The tour of audience knowledge I provide here begins in the 1930’s, at the dawn of audience research, where I draw on historical evidence gathered by other researchers. I then show how these new technical discourses began to gain ground and reshape Hollywood’s business model in the decades following World War II. As I show below, many important discursive shifts occurred during crisis periods in the film industry, those years when profits fell off, box office receipts declined, and theaters closed. This turmoil frequently opened debate within the industry regarding what was causing the slip in profits and what to do about it. These periods proved to be fertile ground for new discursive turns in how audiences were talked about and understood. I selected two of these crisis periods for special attention, as they proved to be formative ones in the 50 creation of new audience categories. These include the early 1950’s, when the industry was confronting what was known as “the lost audience.” A consensus eventually emerged that much of the decline in moviegoing audiences was due to the introduction of television, although this now dominant explanation was one that was debated at the time. A second crisis period was the early 1970’s, when the movie industry was again facing a box office slump. This time, the slump could not be blamed on any new technology. Instead, leading figures in the film industry wondered to what extent Hollywood was simply out of touch with its audience, given the rise of the youth counterculture. Audience Knowledge in Hollywood’s Classic Era For moviemakers working in Hollywood’s “classic” era (i.e., between the end of the silent era and World War II), intuition was a dominant mode of justifying which pictures to make. Collective understandings of what would “work” with audiences held sway, and the tastes of the motion picture audience at large were constructed in the Hollywood milieu—largely without the use of scientific audience research. In an early sociological analysis of Hollywood, Leo C. Rosten characterized the decision-making environment of moviemakers as follows: “Moods rule Hollywood, not data. Financial statements, business indices, even box office returns are used, when they are used at all, not to guide changes in method and policy, but as metaphorical whips on energy and expectations” (Rosten 1941: 253-254). In his work, Rosten characterized Hollywood as a world of egos in which objective financial data was used only peripherally, and where objective data about movie audiences was largely absent from the calculus that moviemakers used to make decisions about movies. Given this, he saw the Hollywood 51 producer as needing to have a limited, though vital, set of skills: “The producer should possess the ability to recognize ability, the knack of assigning the right creative persons to the right creative spots. He should have a knowledge of audience tastes, a story sense, a businessman’s approach to the costs and mechanics of picture making” (Rosten 1941: 238-239). In addition to having a strong business sense, the successful producer also needed to have a set of instincts about talent, about motion picture stories, and about motion picture audiences. Leo Handel reflected on classic Hollywood in an early handbook of movie marketing written in 1950, recalling that “most motion picture executives were content to let product improvement and sales policies rest on their intuitive insight of what the public wanted, rather than on direct contact with the consumer” (Handel 1950: 4). These “intuitive insights” were an important, albeit intangible ability upon which producers rose and fell in the industry. If this intuitive way of making sense of audiences was the lone repertoire that moviemakers had for talking about audiences in the production of classic Hollywood films, new types of discourses were being introduced that would challenge and alter the nature of moviemakers’ collective intuitive sensibilities about movie audiences. Although a few scattered executives favored drawing on input from audiences provided from psuedo-scientific sources of convenience (e.g. weighing stacks of fan mail to gauge audience appeal around a particular star, evaluating the opinions of theater owners/managers, etc.), there was initially no information about audiences produced in the guise of systematic, scientific audience research. It was not until the work of George Gallup’s Audience Research, Inc. (ARI) was introduced into Hollywood during the 1930’s that moviemakers gained an additional repertoire for talking about audiences— 52 one that spoke of moviegoers as sources of data with scientifically measurable characteristics, habits, and tastes. The proliferation of this new kind of technical knowledge about audiences over the next several decades ultimately challenged several core intuitively-based assumptions about who the motion picture audience was, and thus began to draw a more complex map of American moviegoers. Moviemakers Perceive Their Audience’s Social Class When ARI’s work first began in the 1930’s, movie executives had several assumptions about movie audiences that are very different than the ones that moviemakers in Hollywood commonly hold today. One way that moviemakers looked at audience appeal differently was that “prestige” pictures were still conceived of as heavily profitable ventures. These were pictures that were conceived of as highly theatrical epics that were thought to attract people of higher socioeconomic background, a group initially thought to be key to the commercial success of movies geared toward mass audiences. Studios attempted to organize the exhibition of such pictures in a way that was modeled on the experience of stage theater, so that seeing a movie like Gone With the Wind (1936) in some theaters would involve more formal attire, an intermission, and more ornate surroundings (Balio 1993). It was this perceived imperative to draw in higher class patrons that frequently led studio executives to mine Broadway for motion picture ideas, and led moviemakers to push the idea of film as a legitimate art form by encouraging an entire critical apparatus of awards, critics, and film schools (Baumann 2007). According to Susan Ohmer’s history of audience research in this era of Hollywood moviemaking, some of ARI’s first audience research in 1941 argued that a 53 large segment of the motion picture audience was in fact not of high socioeconomic background—a finding that went against the orientation of many studio executives at the time. In fact, the survey (which was conducted for RKO) reported that those coded as “lower middle class,” “poor,” or “on relief” accounted for 47% of Hollywood’s box office income (Ohmer 2006: 129). David Ogilvy, who was in charge of interpreting the findings of ARI, argued forcefully to RKO’s executives that Hollywood should awaken to the reality of its audience’s socioeconomic background, and stop behaving as though the tastes of studio executives were also the tastes of moviegoers. As Ogilvy put it in one report to executives, “If Hollywood producers could be persuaded to spend one week every year in the company of our interviewers, exposing their production plans to coal miners and steel workers, shopgirls and school children, plumbers and builders, WPA workers and policemen, colored and white, they would never be surprised or incredulous” at ARI’s findings (ibid.). Overall, Ogilvy drew on this “common man” imagery to equate moviegoers of lower socioeconomic background with the “mass audience” more generally. In a report to executives entitled “Increasing Profits Through Continuous Audience Research,” he wrote: When films cost fortunes and distribution has only a few main channels, it is imperative that every feature should appeal to the largest possible number of theatergoers. At present there is no room for pictures which appeal to minorities. For better or for worse, the industry is geared to the majority. If the majority does not want the picture, the investment is jeopardized. It therefore becomes of paramount importance for the producer to have a clear picture of the majority. The more he knows about ordinary, average theatergoers, the less likely he is to make costly mistakes in estimating their reactions. (quoted in Ohmer 2006: 130-131) As with any form of scientific research, conclusions drawn do not simply spring from the “facts” themselves, but depend on the interpretations of individuals. Although 54 Ogilvy was steadfast in preaching that Hollywood should orient itself more around its less affluent patrons, it is also worth noting that a majority of moviegoers (53%, taking ARI’s numbers at face value) were not “lower middle-class,” “poor,” or “on relief.” It was also a leap to assume that this group had significant, meaningful differences in their tastes in movies than moviegoers with more income. Even given the pretense of objectivity in Ogilvy’s analysis, there was initial guesswork that went into characterizing groups along lines of social class. 3 Moviemakers’ Early Perceptions of Their Audience’s Age The ways in which moviemakers talked about “youth” also underwent a transition in the early days of audience research. Based again on survey data gathered from the population at large, ARI uncovered a new “fact”: 34% of all movie tickets were bought by those under age 20, and 19-year-olds bought more tickets than any other single age (Ohmer 2006: 132). This also flew in the face of conventional thinking within the film industry at the time, which tended to view the young moviegoing population as insignificant and, more generally, looked to age differences only peripherally to help explain movie attendance (if moviemakers looked to age at all). As Ogilvy wrote in one report, “Until Audience Research, Inc. revealed the fact that 34 percent of all tickets are purchased by persons under twenty years old, there had been a tendency to dismiss the adolescent market as chicken feed” (ibid.). Looking at patterns in movie attendance by age more generally, ARI’s early statistics showed that people under 30 bought 65% of all 3 Ohmer also discusses a case of a clear disjuncture between ARI’s survey data and actual interpretations as to what this data meant. Some early surveys on the relationship between sex and moviegoing uncovered that most moviegoers were in fact women, even while Ogilvy continued to emphasize the importance of men. 55 film tickets, and that, of those ticket sales, people under 20 accounted for half. Ogilvy’s interpretations of these statistics in reports to studio executives repeatedly emphasized the importance of making films for teenagers, much as they had emphasized the importance of appealing to lower socioeconomic groups. For the first time, then, ARI’s studies provided a picture of the “average” audience: young people (especially teenagers) and moviegoers lower on the socioeconomic ladder. From that point forward, ARI began oversampling younger moviegoers in its survey-based studies, preordaining the extent to which youth would matter in interpretations of data on moviegoing audiences. As was the case with researchers’ early interpretations of social class and movie attendance, how data on the age of moviegoers was interpreted at the dawn of audience research raises as many questions as it does provide definitive answers. Assuming ARI’s data was accurate, other interpretations can be imagined. If 65% of all tickets were indeed sold to people under 30, then 35% were sold to people over 30. Might this group of people over 30 just as well have constituted a meaningful demographic group with discernable tastes as the 33% of moviegoers who were under 20? And why were people ranging in age from 20 to 30 lumped together with adolescents? Might this group be more similar in their movie tastes to those 30 and over? Using ARI’s audience data, a different way of looking at Hollywood’s age equation in the early 1940’s could be to recognize that approximately two thirds of the audience was over age 20, past the peculiar, idiosyncratic stage in the life course marked as adolescence. Ohmer’s telling history reveals how key assumptions were made in early audience research that would profoundly influence precisely how the American moviegoing public was carved up and given identities that moviemakers saw as meaningful. 56 All Shook Up: The 1950’s and the ‘Lost Audience’ In his 1950 report on movie audiences and the techniques used to measure them, Leo Handel characterized the industry’s outlook over the coming years as follows: Social and economic trends show that the motion picture audience in the United States will probably experience a steady growth. A study by the 20 th Century Fund indicated an enlargement of the domestic market for films because of growth of population, continued reduction of the work week, wider distribution of national income, and changes in the spending pattern of the average family, with a greater proportion of expenditures going into recreation. (Handel 1950: 96-97) Soon after this optimistic outlook was published, however, the Hollywood film industry found itself in the midst of a crisis, where it faced declining attendance, declining box office returns, and declining profits. Within the industry, this was eventually called the problem of “the lost audience.” The audience for Hollywood movies appeared to be shrinking, and a debate began as to what was causing it. One explanation was that the rise of television was displacing the leisure time that “lost” moviegoers formerly spent going to the movies. This was an explanation that eventually became a consensus view, and it remains an explanation I heard during the course of my fieldwork when this time period was mentioned (although, in point of fact, film attendance had begun to fall off precipitously immediately after World War II, a few years before television started becoming widely diffused). A second explanation was that movies themselves were lacking, and simply failed to resonate with audiences as they once had. Several reports from Variety expressed this sentiment. As one article that reported on a survey that had recently been conducted with theater owners put it, “Theater managers checked in a cross-country survey are convinced 57 that poor product rather than public poverty, and telegraphed plots rather than television, are to blame for the current national box office decline” (Kaplan 1951). Another article framed the blame in terms of there being glut of unsurprising, “mediocre” films on the market. Top exhibs state that the conclusion is obvious: The public won’t buy mediocrity. There’s the high-cost-of-living angle, which means the customers are more selective in spending their limited spare dollars, plus the attractiveness of television in comparison with so-so pix which mean an immediate cash outlay. (“Gotta Be Smash Hits…” 1951) In the midst of this crisis, the tone of the reporting in Variety sometimes portrayed an industry in the throes of desperation, searching for some answer to the downturn. One article represented the industry’s mood as follows: “The slogan currently being voiced is try anything new or old, for that matter, in an effort to find some panacea for soggy grosses” (“Try Anything as B.O. Panacea…” 1950). Drawing on longstanding intuitive discourses and nascent technical discourses about audience tastes, this environment spurred talk about what a cure for the box office doldrums might entail. Interpretations in the intuitive vein saw a need for films that were understood as “more entertaining.” The 1950 survey of exhibitors that Variety reported on testified how movie theater owners saw their audiences as wanting “something new” out of films in the 1950’s, and that these films should focus on “escapism.” According to this interpretation, Hollywood had been making too many films that were too “serious” to be entertaining. As one exhibitor put it, “If people want depressing stuff, why do they go to baseball, basketball and football games—or sit home and play Canasta? Hollywood has been asleep long enough. The business is here—100,000,000 people looking for entertainment. Hollywood should wake up and give it to them. Bring out more comedies and good musicals. Avoid 58 Figure 2: “Optimism is Sweeping the Industry” A 1951 Variety advertisement from MGM, touting reasons for a positive outlook in the midst of the box office slump of the early 1950’s. 59 depressing stuff” (Kaplan 1951). Other articles also spoke to an imperative to find “escapist” fare. “Newly added to the taboo list of subject matter,” one article reported, are heavy-handed dramatic subjects or anything of a downbeat or depressing nature. The reason is the obvious one, of course—the b.o. Numerous top-ranking pix in these categories in recent months have proved box-office fizzles. They’ve received good notices and have been well-liked by audiences that have seen them, but it’s been impossible to get sufficient audiences in. The public is apparently scared off by the subject matter. (Golden 1951) Embedded in this collective intuition that producers used to interpret the success and failure of movies was a theory of human nature: people in general do not find serious themes entertaining. Instead, the driving reason that people go to movies is to escape and feel good. Other possible ways of conceptualizing “entertainment” were excluded from this definition. At the movies, audiences might be riveted on the story or imagine the world as they never had before, but these were not the core characteristics that made movies “entertaining.” Nor were they discussed as the core elements lacking from the movies in the poor-performing movies of the 1950’s. Entertainment, above all, had to be light and unchallenging (emotionally, morally, or politically). As I discuss in Chapter 3, this understanding of “entertainment” has carried through to the present-era film industry, where films that deal too much in “seriousness” are deemed unsellable to wide audiences. If these intuitively based discourses about audiences showed continuity in how moviemakers interpreted box office successes and failures, then technical discourses provided a way of interpreting audiences that was still relatively new. Various projects using the language of scientific audience research were underway, all aimed at cracking the problem of the “lost audience.” In 1950, Gallup’s Audience Research, Inc. was 60 continuing its work by undertaking a nationwide market study to “not only find out who is going to the movies today, but to investigate various factors which may account for the fact that too many persons are staying away” (“Biz Perks…” 1950). Theater owner associations were also doing their own research to find out why people were “staying away from pix.” Western Pennsylvania Allied as well as Associated Theatre Owners of Indiana both undertook their own local studies, posing survey questions like, “If you seldom or never go to movies, what is your main objection?” directly to customers. Western Pennsylvania Allied was “asking its members why they should depend on professional pollsters… to give them an answer to the current box office doldrums” (“Drive to ‘Poll Your Own Theater’” 1951). Scientific audience research was being offered as a way to address the problem of the “lost audience,” and being received in the industry as a viable strategy. Emergent technical discourses about audiences ultimately encouraged an increasing awareness and more intense focus on audience segments. By conceiving of the motion picture audience in terms of demographic characteristics, moviemakers might create a typology of what types of moviegoers (on average) had been “lost” and what types of moviegoers still went to movies. If this sort of knowledge was on hand, moviemakers could formulate a strategy for recovering lost ticket sales. Such a strategy could take at least a couple of forms: it could focus on trying to bring “lost” audiences back into theaters, or it could focus on more intensely appealing to those groups who appeared to still be going to movies. Some of the “lost” groups that Hollywood spoke about in the early 1950’s included older moviegoers, more educated moviegoers, and women. And, at times, it 61 seemed debatable which of these groups was most important. An ad for McCall’s magazine published in a 1951 issue of Variety, for example, provides a testimonial letter from RKO executive Jerry Wald on the significance of adult women in comprising the “lost audience” (see Figure 2). “The “lost” audience in the motion picture industry today is not the television fans, nor the intellectuals, nor the over-30 group, as has been suggested.” Before endorsing McCall’s as a source for stories “that appeal primarily to women,” Wald argued that more pictures should be made that appeal to “the woman who formerly went shopping with her neighbor, then took in a matinee” and the woman “who dragged Pop and Junior out of the house once a week to see the double bill at the Bijou” (Variety 1951: 19). Such a strategy emphasized bringing lost moviegoers back into theaters. Eventually, however, these “lost” segments started to become deemphasized in discussions of how to recover from the box office slump. “Lost” appeared to mean lost forever, and the segments that remained now comprised Hollywood’s newly and more narrowly defined core audience—namely, the “youth” audience. The idea that “youth” represented a relatively large segment of Hollywood’s audience was not a new one, but now this idea started to become a more prominent and enduring feature in Hollywood’s model of its audiences. In 1950, Leo Handel had summarized the state of the field on the youth market. “All surveys,” he wrote, “point to the fact that young people attend more frequently than older people. Many of the older persons, as pointed out before, did not have the opportunity to become motion picture fans when they were young, simply because the film industry itself is young.” Thus, there was nothing inherent in being older that caused one to go to the movies less. Measured differences in how movies 62 Figure 3: The “lost audience” Variety advertisement from 1950. McCall’s magazine weighs in on the problem of “the lost audience,” publishing a letter from an RKO executive that argues that more pictures should be made that appeal to “the woman who formerly went shopping with her neighbor, then took in a matinee.” 63 appealed to people of different age groups simply had to do with the fact that older people in 1950 had not experienced movies as much during their formative years. Later, this explanation as to why older people did not currently go to movies as frequently would be dropped; younger people would simply go to movies more frequently because their lifestyles supported it and they simply liked movies more. The “lost audience” of the early 1950’s thus resulted in more talk within the industry that drew on scientific audience research as well as a heightened focus differentiated audience segments, with “youth” forming a key segment in the industry’s changed understanding of who movie audiences were. Youth audiences were being spoken about as fundamentally different that other audiences, a segment whose generational sensibilities stood in opposition to older ones. As one Variety article worded it, “…a new audience has grown up, dominated by youngsters with new ideas about romance and beauty, and their wants should be catered to by Hollywood” (Kaplan 1950). In the 1970’s, the category of “youth” would be further elaborated upon and new audience categories would spring from it. The early 1950’s, however, marked the last moment in the film industry’s history when older moviegoers (i.e., those over 30) would unquestioningly be talked about as members of Hollywood’s core audience. The Times, They Are a Changin’: The 1960’s and 1970’s Counterculture and Perceiving the “Youth” Market In the early 1970’s, the film industry perceived that it had yet another crisis on its hands. The 1960’s had proven to be a lackluster decade in terms of box office revenues, and theater attendance had a continued its now slowing decline. In addition, the industry 64 was making and trying to sell pictures that some in the industry saw as out of step with current social and cultural currents (Biskind 1998). Hollywood’s lighthearted musicals seemed mismatched with the anti-War movement and the various countercultural movements seen on college campuses. In this climate, a great deal of debate occurred over the meaning of “youth” and its significance for the film industry. Who was this new generation? And what sorts of films did they want? During the early 1970’s, the notion of “youth” as a distinct segment that was opposed to the tastes older audiences became thoroughly normalized in the industry’s map of its audiences. Herein, there was just something about being young that made young moviegoers irreconcilably different from their older counterparts. This realization (in the industry’s eyes) formed the basis for map of audience appeal that used age as a primary marker of difference. The Youth Market and the Birth of Blockbuster Summers The rise of “youth” as a distinct demographic group as well as its rise in strategic importance for the film industry can be measured in part by the birth of the summer blockbuster season. As difficult as it is to imagine now, the summer was once understood as a relatively slow season within the Hollywood film industry. One Variety article from 1950, for instance, reported how the major studios were scheduled to release a remarkably large number of pictures in June, July, and August of that year—in what the writer called “the usually slack summer months” (“Top Pix…” 1950). And during the film industry’s apparent recession of the early 1950’s, one article described how “the summer hot spell threatens more theatre shutterings than at any time since the war ended” 65 (“Try Anything as B.O. Panacea…” 1950). The peak box office time for many years was understood to be late December and early January, and the studios formulated their strategies and release schedules accordingly. Releasing big movies to coincide with the Holiday season not only took advantage of the fact that children were on break from school, but also took advantage of the fact that many adults would have more time off from work and have more leisure time available to them. As late as 1970, just when more sophisticated computerized data on box office receipts was coming into use, one article in Variety stated that, “It comes as no news in the film trade that the supreme box office peak period of any season comes with the Christmas-New Year holidays” (“Bigger X-mas…” 1970). Figure 4: Understandings of seasonal box office returns, c. 1970 A graph of average box office returns by season for the first such calculations done in the industry, for the years 1968 and 1969 (published in Variety in 1970). For these calculations, the period from Christmas through New Years day was still seen as the highest grossing season. 66 Figure 5: Variety’s box office barometer, c. 1983 A “box office barometer” graph from a 1983 issue of Variety. The current year’s performance was now compared to a “13-year average” of box office performance. Figure 6: Variety’s box office barometer, c. 1989 A “box office barometer” graph from a 1989 issue of Variety. The current year’s performance was now simply compared to the “expected” level for that season. 67 By the 1970’s, the rising importance of the “youth” market and the declining importance of older moviegoers helped usher in a new era in how the film industry conceived of summertime movies. Releasing movies during the months of June, July, and August is a strategy designed to cater specifically to moviegoers under age 25 (and under 18 especially), many of whom have a great deal of newfound leisure time given that they are on summer vacation. The ascendancy of “youth” coincided with changes in the technology used to measure box office returns as well as landmark summer blockbuster successes (e.g., 1975’s Jaws). Within the industry, all of these forces coalesced to bring about a new understanding of what kinds of films (and what kinds of profits) were possible during the summer months. When a large scale survey of movie theaters was used to estimate total box office returns by season in the late 1960’s, it made the idea of “average” seasonal box office possible for the first time. Using what was then a state-of-the-art computer (the IBM 360), the seasonal box office averages provided the first technical artifact lending credence to the notion that there were natural ebbs and flows in box office returns over the course of a given year (Silverman 1970). Of course, the baseline “typical” year in this case was 1969, and successive years would always be compared to the years before it. Whether or not 1969 or the succeeding years were in fact a “typical” years was less important than the fact that objective comparisons of the film industry’s performance over time could now be made. Such averages appeared immune from the idiosyncrasies of any one studio’s individual performance, and, after a few years, immune from the peculiarities of any particular year. Over the years, the baseline for computing averages included more and more years, and the meaning of these averages changed from being a 68 novel piece of information to representing the “expected” level of box office returns for any particular month. Thus, such averages of box office performance eventually provided the basis for the principle that there were natural seasonal differences in box office appeal—and that the summer months were the strongest months. Yet, despite the increasing sophistication of computers used to calculate seasonal box office averages and despite the fact that summers did in fact come to represent a time of huge blockbusters, there is no escaping the fact that this new knowledge helped produce the very truth about the summer months that it proclaimed. Computerized box office measurement technologies began to fully mature just about the time that the industry was sensing heightened anxieties about appealing to a highly visible “youth” counterculture that was thought to be different than moviegoers in general. Thus, more movies would be targeted specifically at this group in what was thought to be its season of choice, causing the summer months to experience year over year growth (in revenue and perceived importance), and, after several years of amplification, eventually giving the youth-summer-blockbuster equation the appearance of natural law. To fully understand how the early 1970’s gave birth to the blockbuster as we understand it today, we must return to the talk and debate about the “youth” audience that occurred within the film industry during those years. The Oppositional “Youth” Audience The film industry’s sense of crisis in the early 1970’s resembled the crisis in the early 1950’s insofar that there was still a lingering allusion to a “lost” or “missing” public. One Variety article from 1970 reported that the National Association of Theatre 69 Owners and the Motion Picture Association of America were undertaking an effort “looking to investigate the composition of theatre audiences—specifically, the reasons why the public in part is buying and in greater part is not buying tickets” (“NATO- MPAA Eye… 1970). The hope of bringing various “lost” moviegoing segments back into theaters may have still been alive, but these audiences were already conceived of as different from young audiences now considered to be the bread and butter of the film industry’s business. If it was possible to bring older moviegoers back to the movies, Hollywood would have to do it by adding films to its slate that were different than the “youth-oriented” pictures it was currently offering. Among some industry analysts, there was a sense that the industry was on the brink of a sea change. One writer, in a front page Variety article from July, 1970, went so far as to proclaim that “the next 11-12 weeks could be the most fateful in the history of the motion picture business. Seldom, if indeed, ever, will the market have been as crowded with top-budgeted films, ranging the gamut of entertainment interests from the young to the middle aged and older audience melting pot” (Pryor 1970). But if this was an “audience melting pot,” industry interpretations of social and cultural events were starting to cast younger moviegoers as though they were indelibly frozen in their own age-based enclave. No other group helped feed this perception more than radical elements within the counterculture. In 1970, there were worried reports in Variety that a student-led boycott of Hollywood films was being planned to push the industry to reflect students’ generational concerns (“U.S. Youth Threat” 1970). The newspaper also reported on campus culture as it pertained to the film business, and one report quoted at length from Herbert Marcuse’s critical analysis of the mass media (with 70 Marcuse being called the “chief philosopher” of the youthful New Left) (Landry 1970). There were also scattered reports of violent disruptions at movie theaters, as the counterculture hit its radical height. One Variety writer characterized this climate as follows: “Back in the simplistic days beyond recall, say, 10 years ago, bad business at the film theatre box office could be blamed on bad product, television, weather.... Today all these detriments to showmanship yield to updated fears and within recent months, a fear hardly heard since the riots and sit-down strikes of the Depression years: violence in the streets…” (Landry 1970). Although there were fears within the industry of the counterculture’s more radical elements, there was also a much bigger fear that the film industry was out of touch with this peculiar new generation. Given this, there were some signs that studios were attempting to adapt to this new market. One article reported that “the youth market has been for several years the film industry’s most imposing but neglected enigma.” Given this, major studios were now frequently appointing “fulltime scholastic contacts” whose function was to stay on top of youth sensibilities, and that such contacts provided feedback that “looms as prime source of required reading on a film’s potential.” (Spilker 1970). Even Metro-Goldwyn-Mayer’s trademark, a growling Leo the Lion, was temporarily dumped to cater to this new generation that was increasingly being seen as different from other Americans. “‘Nostalgia be damned’ was the theory, and once Leo was sent to the Memory Zoo, in came a new corporate lion, obviously a graduate of the Harvard Business School: sleek, …stylized, and ‘now.’” That said, the new “now” Leo the Lion had a very short life span, and the old one was soon reinstated. Referring to the change in logos, the Variety writer said, “That may have done wonders for corporate 71 egos, but for the general public and heartland exhibitors, it just didn’t wash.” (“Banished Leo the Lion…” 1970). Talk about what sorts of films “youth” liked and what sorts of films were “now” films framed many Variety reports published in the first years of the 1970’s. Here, “youth” was frequently associated with what were then referred to as “message” pictures (i.e., films that had some critical commentary to give on contemporary politics or American society) and “now” pictures (films that were roughly understood as having something to do with rock n’ roll, drugs, and the sexual revolution). Under this thinking, studios could pursue either the “youth” audience with its “now” sensibilities, or it could pursue other demographic groups (older moviegoers, “families,” etc.). But they could not hope to appeal to both sets of moviegoers with the same film. And although the highly visible youth culture dominated many headlines, others in the industry touted the virtues of going after other audiences with other sensibilities. Speaking of the rise of “message” pictures, George Seaton, director of 1970’s commercially successful Airport, was reported as having said, “The only problem I see is that this trend could get out of hand and moviemakers start pitching all their pictures at the youth audience. There are millions and millions of people over 30 who still want screen entertainment in the old sense” (“Can’t Let Youthful…” 1970). Another Variety article told of how one Hollywood representative, after touring several Northeastern cities, reported back that people “don’t want pictures with a message—at least with a message that’s driven home with a sledgehammer” (Scott 1970). At the corporate level, some studio leaders were commenting on how their strategies were departures from youth-oriented strategies. For instance, one article reported that “Columbia Pictures has become non-conformist among 72 the film companies, in that it is not trying to place its product exclusively with the youth market.” In addition, another report told how Irving Ludwig, who was president of Buena Vista (the distribution arm of Walt Disney Productions in 1970), “insists that the family trade is ‘the biggest group of audience we have.’ And while the new permissive or ‘now’ pictures make it in the first runs in some cases, they eventually ‘taper off,’ as he clocks the turnstiles” (“Big Coin Aud Still ‘Family’” 1970). “Youth” as “Film Buffs”: Anticipating the “Independent” Audience “Independent” filmmakers that emerged in the 1990’s initially conceived of themselves as the antithesis to “Hollywood” films, and could trace their roots to the underground cinemas of previous generations of filmmakers (Levy 1999). In the early 1970’s, the Hollywood film industry was recognizing a kind of proto-“independent” movie audience as existing within the “youth” audience whom it was giving so much attention. In Variety articles from this time period, these young moviegoers were called “film buffs.” They were young audiences who had now grown up on movies, perhaps attended college classes on film, and had grown to be more sophisticated, critically savvy patrons of movies than the generations that preceded them. One wry commentator on this new segment of filmgoers caricatured this new “species” in his observations of moviegoing in New York City: 73 Figure 7: “Film buffs” advertisement, c. 1970 An advertisement in Variety taken out by Paramount pictures in 1970. Motion picture studios sought to understand the counterculture of the 1960’s and 1970’s as an important element of the now all-important “youth” demographic. 74 Figure 8: Viva Max! advertisement, c. 1970 An advertisement for Viva Max! from 1970. Twin anxieties were voiced in the film industry during the early 1970’s: being able to cater to the “youth” audience while not alienating “families.” 75 He was a New York cineaste—a special breed of bird, native to the Eastern seaboard, who has come into prominence over the last decade. Bookstores have been remodeled to suit him, and the book industry has revamped its lists to cater to him. Third Avenue, once studded with bars, is now a maze of art theatres, piled literally one on top of the other, to feed his voracious appetite…. As likely as not a product of the estimated 2000 film courses now given in colleges, YMCA’s, and other cultural bastions of the nation…. (Bruno 1970) Film buffs were seen as fitting with a countercultural emphasis on challenging formal, stylistic, and thematic conventions. In the early 1970’s, this contrasted with more “mainstream” fare of the kind that Hollywood “used to” produce—films that might be financially successful but were unchallenging from a critical standpoint. Referencing two pictures that were coded as “not youth oriented” and not of the kind that would appeal to “film buffs,” one Variety writer put it as follows: “The Adventurers and Airport are clearly throwbacks to the days of the Hollywood dream factory so recently pronounced forever extinct. Critical assessment of quality aside, these two are drawing that lately neglected older generation of filmgoers” (Verrill 1970a). Such a sentiment represented a conflict drawn along lines of generational sensibilities. Some reporters for Variety who followed the youth counterculture remarked how younger sensibilities were moving further from the mainstream and more into avant-garde themes. These writers saw youth as having an intensifying interest in “underground” cinema and all that it represented for the industry. One writer spoke about the connection between “youth” and underground cinema, saying that “today’s young student filmmakers, together with many vet underground experimenters, are becoming increasingly obsessed with images of death and filmic projections into a bleak future for the U.S.” (Verrill 1970b). The implication 76 was that this focus on “bleakness” placed this increasingly “youth”-driven movement outside the mainstream interests of everyday, middle Americans. The fact that much of the anti-War movement and the youth counterculture was centered on college campuses, with their burgeoning film schools, may have led some Variety observers, in the final analysis, to see “youth”-oriented films as belonging to a niche audience that was differentiated by its love of “messages” and its higher brow level. In the early 1970’s, younger, more educated filmgoers became associated with “message” pictures, whereas before industry observers considered the possibility that the public at large was becoming more interested in “messages” and “now” pictures. As one reporter put it, “The films of contemporary alienation have been supported almost entirely by the better educated, upper middle class kids. One of the great seeds of resentment in the silent majority of blue-collar workers and ‘hard hats’ has been the thought, ‘What the hell do these kids have to complain about?’” What this “silent majority” wanted, the writer said, was “lighter entertainment..., escapism, romance, and comedy….” (Setlowe 1970). Thus, the division between these two audiences, such as it was, began to take its current form: one audience (a slim minority) was seen as wanting political, artistic, and cultural statements in their rejection of mainstream Hollywood fare, while the less privileged majority wanted what was broadly understood as “entertaining” (escapism, romance, and comedy). By a little more than a decade later, in 1983, it was possible to talk about these two kinds of audiences as two objectively real markets. “Message” pictures became “artsy” films for the studios’ new “classics” divisions, which existed in opposition to more mainstream “commercial” fare (Klain 1983). When this last discursive shift occurred, the contemporary division of moviegoers into 77 “commercial” and “artistic” or “independent” audiences had been born (see Chapter 3 for further discussion of this division as it is manifested today). Hollywood’s Map of Audiences and Its Doubters By the mid-1970’s, the Hollywood film industry had divided its audience in several fundamental ways. These divisions were arrived at through fits and starts, mostly during periods of financial crisis in the industry’s history. The various ways that the film industry now divides and categorizes its audiences amount to a collective knowledge of audience appeal that sees particular kinds of movies as belonging to particular kinds of audiences. As discussed above, talk in the film industry had come to foreground two important lines of difference within the once general population of moviegoers by the mid-1970’s. First, this included a stark conception of how young moviegoers differed from older moviegoers. Defining “young” moviegoers as people 25 and under, defining “older” moviegoers as people over 25, and seeing these two groups as irreconcilably different was a gradual process that began with early interpretations of scientific audience research in the 1930’s and culminated with the financial successes of the blockbuster audience in the 1970’s. A second key division that the film industry began to see were differences between what was eventually called the “arthouse” audience and what was eventually labeled as more “commercial” audience. In response to the particular social context of the youth counterculture of 1960’s and 1970’s, talk within the film industry began to speak of differences between a more “mainstream” audience that liked “entertaining,” escapist themes and a more avant-garde audience that liked more challenging, 78 “message”-oriented themes. Although “underground” cinema had already existed in the United States for a very long time by the 1970’s, the themes popular in the underground now were seen as speaking to a sizable market of alienated, upper middle class young people belonging to the counterculture. Figure 9: 1975’s blockbuster summer A cover of Variety during the summer of 1975, the year Jaws was a box office phenomenon. With the birth of the modern blockbuster in the 1970’s, the studios shifted to a more intense focus on the “youth” audience and the summer months. Once in place, these ways of constructing audiences served as powerful ways of justifying which sorts of film ideas belonged where. In succeeding years, some industry analysts would occasionally question the utility of adhering to such divisions— particularly when it came to Hollywood’s intense focus on “young” audiences. One Variety reporter writing in 1986, for example, noted how that year’s comedy films were 79 violating Hollywood’s assumption that young moviegoers were an enclave all to themselves. “While it’s true summertime is traditionally funnytime at the movies, observers feel summer 1986 has so far been notable for the broad demographic appeal of the straight comedies and comedy hybrids…. Industry insiders note the ability of this summers boffo laughers to tickle tastes across age-group lines contrasts sharply with the fizzle last summer of strictly teen-oriented fare” (Gold 1986). Seven years later (in 1993), another article described how Hollywood’s older demographics were potentially being underserved. “While movie auds are still composed largely of 16-30-year-olds, the size of the over-40 audience has doubled in the last decade to represent 30% of ticket sales. And the percentage of the population that says it never goes to the movies has shrunk from a high of almost 45% to less than 30%.” Carl Patrick, the chairman of Carmike Cinemas who was interviewed for the article, commented on these facts. “‘The truth is we hardly touch the 35-million-40-million retired people in this country…. It’s a huge audience with a desire to go, and the money to spend, on films like Fried Green Tomatoes. There just aren’t enough pictures for all seasons and for all ages’” (Klady 1993). Occasional doubts withstanding, the Hollywood film industry has mapped the moviegoing audience in ways that continue to frame the decision-making environment in the film industry’s disperse production sites. Changing Audience Knowledge and Research Organizations How did Hollywood’s understanding of American moviegoers come to shift so dramatically between the 1950’s and the 1970’s? That is, how did the industry stop talking about the possibility of appealing the “moviegoing public” at large and eventually 80 see this public as irreconcilably divided into segments? I see the introduction and proliferation of media research firms as a prime catalyst for these changes. Beginning with early studies conducted by George Gallup’s Audience Research, Inc. through the present day work of Nielsen NRG, MarketCast, OTX, as well as the studios’ own research departments, a fundamentally new kind of technical knowledge gained prominence in Hollywood during the mid-twentieth century. Through the entrepreneurial work of these organizations, who brought with them computerized measurement technologies as well as marketing experts trained in selling consumer products, the industry first became aware of average differences in audience behavior by demographic characteristics. These “objective” facts (as they were interpreted) helped reify audience categories, eventually setting in motion a cycle in which movies were continually targeted at segments of moviegoers who were already thought to go to particular kinds of movies, thereby compounding the idea that those segments in fact had real, objective tastes. But, it was not so much that new technical discourses displaced older intuitive discourses for talking about audiences. Rather, intuitive discourses (to some extent) incorporated technically-informed ideas, reshaping the industry’s collective intuition about who moviegoers were and what sorts of films they preferred. Importantly, collective intuitive discourses began to reflect Hollywood’s increasing division of its audiences. In the early 1950’s, intuitive discourses spoke about movies as either “entertaining” or “not entertaining.” Accordingly, successful movies provided an adequate level of escapism and lightheartedness to appeal to moviegoers at large. By the 1980’s, core intuitive descriptors had shifted to focus on whether movies were 81 “commercial” or “artsy.” These new terms referred to films with particular types of themes or looks, but they also implied that different segments of the audience went to these different types of movies. Once movies were spoken about in this way, both intuitive and technical discourses continued to be reproduced in the varied organizational contexts of Hollywood film production. 82 Chapter 3: Intuitive Knowledge in Reproduction The problem is, people really don’t do the kind of analytical work when they’re conceiving movies that you might think they do, or what they would like you to believe they actually do. There is an inordinate amount of seat-of-the pants knowledge of the marketplace that people rely on. It’s not empirical. —Former studio head, in an interview Every few years in Hollywood’s collective discussion of itself, someone proposes a new scheme for objectively predicting box office success. In one recent iteration of this proposition, a company called Epagogix has been developing a large computerized algorithm for predicting why some movies happen to “hit” with audiences and why others are flops. Epagogix’s algorithm is based on what is known as an “artificial neural network,” which inputs thousands of variables along which narrative and plot structures of movies might vary, and then averages the relative effects of those variables over hundreds of movies (Gladwell 2006). Developers of such algorithms are typically very upbeat about the eventual application of such an objective formula to the Hollywood movie business. If ever successful, Epagogix’s artificial neural networking could fundamentally change the way movies are made, objectifying what has been thought of as a subjective process, and make Epagogix a great deal of money. Others working close to the industry maintain a deep skepticism over the possibility of ever making the process of choosing movies a truly objectifiably accurate one. Perhaps the most oft-quoted figure on the subject of choosing hits is William Goldman, who famously wrote in his Adventures in the Screentrade (1983) that, “Nobody knows anything.” Some academic researchers, too, even suggest that attempting to map which movies will hit and which movies will not hit involves 83 something approaching a chaotic, random pattern. There are, according to this interpretation, simply too many moving parts involved in selecting movie ideas, making movies as they were intended to be made, and accounting for what is going on in Hollywood’s various markets to ever be able to predict in advance which movies will sell. In this spirit, David Picker, a former studio executive, is quoted in Goldman’s book as making the following reflection on his track record in picking movies: “If I had to say yes to all the projects I turned down, and no to all the ones I took, it would have worked out about the same.” One former studio executive who taught a class also attested to how frustratingly random the art of picking movies seemed. He recalled one point during his tenure as a top executive when the studio’s private financial consultants showed him and his fellow executives how the studio would be better off simply investing the company’s dollars in U.S. Treasury bonds, as this would offer a more profitable option than making movies. Given their recent track record, the studio’s executives had underperformed in their decisions regarding which movies to make. These rival visions of how movies might be selected for production underscore the extent to which prediction remains a central concern for those working in the movie business. “The audience” looms “out there” beyond the studio lots, office towers, and gated celebrity homes that make up Hollywood. Given this hard reality, how do those involved with various aspects of the motion picture business think they are making a movie that will have audience appeal? So far, no algorithm has yet emerged that moviemakers use to replicate movie success. The myriad decisions surrounding the production of movies—from the basic stories involved, to who should play which parts, to how much money to put in the budget, to whether particular films should be made at 84 all—are all decisions that are still necessarily made by people who state that they are simply making their best guesses. In this chapter, I take up the issue of how screenwriters, script analysts, producers, talent agents, and studio executives produce and utilize an intuitive knowledge about movie audiences as they go about the business of evaluating movie ideas for commercial production. Intuitive knowledge involves discourses about why movies work and why they appeal to audiences, discourses that appeal to a principle of individual inspired judgment. Herein, movie ideas are evaluated on the basis of how they “seem” to the evaluator, as detectors of audience taste. Many in the industry fancy themselves as having highly cultivated gut instincts. Just as some music executives have been described as having a “golden ear” for finding hits, movie executives have long justified their large salaries by arguing that they have a personal gift for choosing which movies will work with audiences (see Ahlkvist & Faulkner 2002 for a parallel case in the music industry). Yet, although it is often based upon the assumption that individual evaluators have particular talents, the structure of intuitive knowledge does not substantially differ according to the individual doing the evaluating. Rather, these discourses operate as a shared culture for evaluating movie ideas. Intuitive knowledge is comprised of various categories for where movie ideas belong. Thus, movies can be “commercial,” “artistic,” or “edgy.” These terms are powerful descriptors for those working in the industry, indicating what movies will look like and containing implicit assumptions about where and to whom they will appeal (see also Guetzgow et al. 2004). Within the industry, these shared cultural discourses become important in collective discussions about movies, often 85 as ways of framing ideas for other players in the industry (see Bielby & Bielby 1994). In this way, intuitive knowledge forms a common language for making legitimate arguments (i.e., “justifications”) about movies. This intuitive knowledge about movie audiences is made and reproduced in the context of organizational realities that coincide with intuitive understandings of how audiences work. Intuitive knowledge is embedded in a different set of lived organizational constraints than is technical knowledge about movie audiences. As discussed in Chapter 4 below, technical knowledge surrounding movie audiences is reproduced in an organizational context where interpersonal reputation depends on notions of individuals’ technical competence. In that case, “scientific” audience research involves very formal, routinized work where instruments and measures are taken as fixed criteria for evaluating films. Such “objective” measures tend to flatten organizational hierarchies between market research firms. Intuitive knowledge, on the other hand, is used much more intensely within organizational contexts where interpersonal reputation depends on informal industry-wide conventions for strategic “impression management.” Here, individuals and firms jockey to be perceived as legitimate players among their peers, according to industry-wide standards for doing “reputation work” (Zafirau 2008). This intuitive knowledge resonates within more hierarchical organizational fields (such as the talent industry), where people can and do rise and fall very quickly—based on the perceptions of their peers. Thus, whereas technical knowledge is experienced as something objective and external to the self, intuitive understandings of the audience can be experienced as a more intimate sense of what works with audiences and what doesn’t. Indeed, intuitive knowledge can be reproduced at the intersection of producers’ personal 86 and professional lives, at the overlapping boundary between moviemakers’ organizational and personal selves, flowing into family and leisure time (Zafirau 2009). In this way, intuitive knowledge works as a common legitimating language among people whose reputations are experienced as tenuous, and who work in an organizational fields that are in flux (Zafirau 2008). In the sections below, I first describe the basic processes through which intuitive knowledge is put into material form in the Hollywood film industry. Every year in Hollywood, tens of thousands of movie ideas are introduced into the system, whether in the form of pitches or in the form of scripts. Most of these ideas are rejected early on in the process. Some languish as unproduced properties of studios and production companies. Then, there remains a tiny minority of movie ideas that make it to the screen. I discuss how this vast sea of movie ideas is initially sorted, drawing on in-depth interviews with those involved in writing and evaluating scripts. Second, I discuss the role that intuitive knowledge plays in this process. Here, I describe how evaluations of movie ideas (in the form of scripts) are articulated using a preexisting language that is populated with meaningful cognitive categories. These various categories for describing films (e.g., “commercial,” “artistic,” “edgy”) convey how a movie would look and feel, and are used to justify why a movie should be made as well as to make arguments about which audiences will like a movie. Finally, I provide a grounded account of how this knowledge is produced within organizational context. Here, I draw on participant observation data at a talent management/ production company that I call Limelight Entertainment to show how intuitive knowledge resonates within a world in which the players involved strategically work to secure their reputations. 87 From Script to Movie: How Movie Ideas are Sorted During my time as a participant observer and intern at Limelight Entertainment, a small but prominent talent firm and production company operating in Hollywood, my day-to-day office duties included opening and sorting mail and photocopying. A vast majority of this mundane work dealt with motion picture and television scripts, which flowed into Limelight by the droves. They came by way of personal courier, DSL, or FedEx, usually bearing the branded logos of Hollywood’s top agencies, production companies, and studios. Often, they came as PDF files, delivered to a talent manager’s assistant’s email inbox. Sometimes, forgotten pieces arrived through the fax machine. The noise and whirr of Limelight’s office frequently involved processing scripts. These screenplays were the material manifestations of ideas for movies, and inscribed in their pages were a writer’s creative energies, the advice of his or her agent and/or manager, and often the suggestions of nameless others at a motion picture studio or production company. No doubt thousands of scripts came through Limelight every year. Scripts that were sent from writers who did not have the backing of an agent or manager were automatically discarded, or else placed in a large pile of unanswered fan mail directed at one of Limelight’s celebrity clients (which was eventually discarded as well). The remaining scripts had to be digested and sorted by someone at Limelight. This was done for a couple of possible purposes: managers might have wanted to know if a screenplay would be appropriate for a client, or perhaps Limelight’s production wing may have wanted to work with a writer and possibly try to produce his or her movie idea. Initial evaluations of this mountain of screenplays involve writing what is known in the industry as “coverage.” Coverage is used in talent firms, production companies, 88 and studios and is typically written by assistants or professional script analysts for executives, agents, or producers who do not have time to read this entire mountain of scripts themselves. Although it varies somewhat from company to company, script coverage typically involves several elements: a short plot synopsis of the screenplay under consideration, the evaluator’s narrative of the movie idea’s strengths and weaknesses, an objective set of “box scores” (or, “the grid”) that are used to rate a film along several criteria (see Table 2), and, finally, the evaluator’s recommendation for what to do with the screenplay (e.g., “pass,” “consider,” and “recommend”). The dozen or so script readers and analysts I interviewed described an extremely selective process where less than 95% of the screenplays received a rating of anything other than “pass.” Script analysts and readers reported that these recommendations were largely taken as final. As one analyst put it, “If it doesn’t get past us, it generally doesn’t get made.” Thus analysts and readers often wield a significant, if underappreciated power in Hollywood’s gatekeeping apparatus. Unless there exists a more direct channel to the agent, producer, or executive who must call the ultimate shot, it is most often a script reader who says either “yes” or “no” to an initial movie idea. When I began writing coverage at Limelight, I was initially intrigued by how each screenplay was the possible seed of a full-fledged movie, with all its potential magic. However, the process of writing and making coverage ultimately becomes very routine and mundane in the world of script readers and analysts. Time pressures impinge on evaluators’ ability to consider each script with a fresh, careful eye. Many readers described going through periods of intense activity, where they would have to read three 89 or more 120 page motion picture scripts a night. As one producer’s assistant responsible for covering scripts put it in an interview: You know, if you’re reading a script and it’s midnight and you’re like, “oh, I have to get through one more script.” Unless it’s really fantastic— like exceptional—you’re not even going to want to work in the development process with it. Most of the readers and analysts I interviewed worked on a piecework basis, where they attempted to read as many scripts as possible to try to maximize their income. They were typically highly educated and articulate, but relatively low-paid, contingent workers in the culture industries of Los Angeles. Independent readers reported being compensated anywhere from $50 to $80 a script, while assistants to producers, agents, and executives read scripts as a regular part of their jobs. Independent readers often had other “gigs” that they did surrounding screenwriting, such as running workshops for screenwriters, consulting independently with screenwriters, or attempting to write screenplays themselves. In the context of their work, where hundreds of scripts could be covered in a year and the known odds of “recommending” or “considering” a script would be very low, script readers and analysts tended to anticipate “passing” on most scripts before they even began reading them. One reader put this as follows: You get a lot of crap. And so you’re just like—your expectations are low. So then you’re like, “ugh!” As soon as it [the script] falters slightly, it’s off of the track. Like nobody wants to give it a chance. Even if it is a good basic story. Thus, evaluation of movie ideas occurs amid circumstances in which stories abound— stories that blur together in their plot similarities, their problems, and their predictable verdicts of “pass.” 90 ________________________________________________________________________ Table 2: A sample of “the grid” used in script coverage ________________________________________________________________________________________________ ________________________________________________________________________________________________ Excellent Very good Good So-so Not good Artistically XX ________________________________________________________________________________________________ Commercial XX ________________________________________________________________________________________________ Premise XX ________________________________________________________________________________________________ Story XX ________________________________________________________________________________________________ Main characters XX ________________________________________________________________________________________________ Minor characters XX ________________________________________________________________________________________________ Dialogue XX ________________________________________________________________________________________________ Visual elements XX ________________________________________________________________________________________________ Title XX ________________________________________________________________________________________________ Recommend: Consider: Pass: Source: Coverage manual used at Limelight, which was borrowed from another talent firm. ________________________________________________________________________________________________ ________________________________________________________________________________________________ Additionally, coverage is written by individuals who are subservient to other, more powerful individuals in the industry, people who some readers described as having over-inflated egos. Given this, some readers and analysts told me of a loathing they had for former or current employers who came off like they had it “all figured out.” In one interview, a full-time analyst recounted an incident in which she and her fellow readers 91 met a new executive that had been hired at the company. She recounted this first meeting with the executive as follows: I’m not going to tell you his name. But he came in, no lie, and said, “hello, my name is ______, and I have an MBA from Columbia.” And I smiled and nodded and said something congratulatory, but I’m thinking, “what an asshole.” For the readers and analysts who described their employers in this light, the arrogance of the executive was sometimes attributed to his young age. Accordingly, they were young and had “something to prove.” Or, even if they were not young, readers told how executives had what they considered to be bad judgment. As one script analyst told me: One of the things you have to realize is that a lot of people working in the industry on an executive level, despite having read a lot of scripts, have no idea what constitutes a good script. And if they were asked to write one, they’d have absolutely no idea. Thus, readers and analysts sometimes reported a tension between their own judgments and their employers’. Readers understood executives to have a different approach, despite their high salaries and their clout in the business world. The power relations between executives and the script analysts and readers under them were occasionally reported to have shaped how judgments about movies were made. Accordingly, most of the time executives were not involved in the initial writing of coverage and usually simply took readers’ evaluations as final. However, occasionally executives were reported to be heavily invested in a movie getting made, and attempted to get coverage writers to portray a movie idea in a particular light. One reader who worked for a production company put it as follows: As a matter of fact, I’ve read scripts that executives have already read and liked. And every once in awhile, an executive will say, “look, this is my vision of the picture, and could you slant your report accordingly?” 92 Because what they’re going to do is they’re going to sit in a meeting and they’re going to push to be able to make this particular project. They’ll say, “in your report, could you emphasize this, this and this.” Thus, coverage always had the pretense of being an unbiased evaluation. But occasionally, factors other than the perceived “quality” of the script itself entered into the writing of coverage. At times, these factors could include an executive’s agenda for a movie, which may have been related in part to the relationships he or she was trying to forge and maintain in the industry as a whole. At other times, negative judgments might be amplified in the final coverage report—even when a reader might judge a script to have some merits. This showed how coverage must contain a clear, resounding “yes” or “no,” even when screenplays seem more ambiguous to readers. One reader I interviewed described how this worked when she wrote coverage. If I’m sitting on the fence, and something is like sort of middle of the road, but what the executive wants me to do is tell them whether should we develop this or not, if I’m going to put a pass on it, then in the writing of the synopsis, there are ways to subtly emphasize what doesn’t work over what does work. And I do a comments section where I evaluate it based on criteria A, B, C and D, basically I come up for a reason why we should buy this or not buy this. You know, so I can slant things. In this way, authors of script coverage knew how to frame their evaluations of scripts in languages that others in the film industry would view as legitimate. If circumstances required it, they spoke of how they could press particular buttons to signal to readers (e.g., executives, producers, agents) that an idea was either a worthy or unworthy movie idea. Overall, then, coverage serves a couple of purposes within the Hollywood film industry. One is the very functional purpose of winnowing down a large number of movie ideas to a very few. Yet, another is to provide a qualitative verbalization of how a 93 movie idea “seems.” In this way, coverage represents an initial moment in which someone answers the questions of whether a movie idea has potential, and what sort of movie the idea could potentially be. Which other movies will this movie be like? Would it play to a wide audience or a very narrow one? Is the movie idea presented in the script a “commercial” one? In this vein, coverage verbalizes arguments that executives then draw upon when interacting with their peers in the industry. Several readers and analysts told me how their bosses simply read from coverage as they spoke on the phone with agents or screenwriters, telling these individuals why their movie ideas were being rejected. Similarly, in the rare instance when a movie is to be made, script coverage can also provide arguments for why a movie idea will work. Once it is written, it can provide an executive, producer, or agent with the sort of “ammunition” they need to make their cases to other high-level decision-makers in the film industry. What is noteworthy is that among the thousands of predictions, estimations, and evaluations that are made in script coverage, this rhetorical “ammunition” is provided by judging screenplays using a finite and repetitive set of terms. In this way, a piece of script coverage represents an artifact that shows the sorts of arguments can be legitimately made about movie ideas. Intuitive Languages in Evaluating Screenplays: How Writing, Writers, and Stories are Judged During the course of my fieldwork, I was able to read examples of coverage that readers and analysts had written. In addition to interviewing people who wrote coverage and the executives who used coverage, I personally had the opportunity to cover several scripts as an intern and participant observer at Limelight. I also attended screenwriting 94 trade conventions where issues of submitting screenplays for consideration at studios and production companies were discussed. In these ways, I was able to go inside the process of how legitimate “intuitive” evaluations of movie ideas are made, and how these evaluations are articulated within the film industry. In the course of this research, I found three ways in which movie ideas are intuitively evaluated. These related to the evaluation of (a) writing quality, (b) writers as people, and (c) story ideas. Although only this last kind of evaluation involves discourses relating to audience appeal (story ideas), the other two are part of the context in which the last kind of judgment is made. First, ideas were evaluated according to the perceived quality of writing itself. This usually involved perceptions of writing mechanics, the coherence of the plot, the plausibility of the characters and the dialogue. Embedded in these kinds of evaluations was the notion that writing screenplays is a craft, one that is to be learned, labored over, and perfected. Thus, although a screenplay may have contained a fundamentally good story idea, it may not have been executed well. Failing to adhere to proper writing conventions could serve as an immediate disqualifier for a screenplay, as one producer’s assistant illustrated to me in an interview. When I asked her what she though made for a “bad” script, she responded: Well, somebody who doesn’t even know how to format a script. You know, if we get it from an agency, usually all that really bad stuff is gone. Um, but, if they can’t format a script correctly, if they have no characterization, it’s just… There’s no storytelling at all, it’s just random to them. They’re just there to be “funny” or manipulative in some other way, and, you know, what is the point? There’s no point, it’s just random writing. To me, that’s just crap. The author was just like, “I had an idea and I wrote it down”— without giving it any thought or working at their craft. Some people are like, “I wrote a script, so I’m a writer.” I’m like, “well, not really.” That doesn’t constitute being a writer, writing a script. 95 In this mode of evaluation, then, writing required effort and discipline. Screenplays had to be formatted correctly. Screenplay authors had to be familiar with basic conventions for structuring a motion picture plot and follow them tightly, with scenes connecting coherently to execute the story. In this light, I heard some writers praised for being very “procedural” writers (or, very logical in the way they executed standard plot arcs for motion pictures). On the other hand, scripts could fall short if they did not connect necessary elements of the plot. As one coverage author stated in coverage she showed me for one movie she passed on, “The script itself is clumsily written, with connective scenes so disjointed that we sometimes find ourselves confused.” Furthermore, scripts had to contain well-developed characters and have dialogue that seemed realistic. Thus, in another piece of coverage, the reader wrote: “While full of heroes and villains, the characters are poorly and perfunctorily motivated and sometimes not motivated at all.” A second way of evaluating movie ideas involved perceptions of who wrote a particular script and whether the author’s experience or traits would allow for a successful movie. Sociologists of art and media production have written a great deal on how constructions of an individual creator’s attributes can be a decisive factor in helping to determine whether an individual is allowed to produce within a particular art world, as well as what kinds of products that producer will be permitted to make. Here, symbolic attributes surrounding social differences can be powerful sorting forces affecting aspiring producers (see Bourdieu 1993). In the world of Hollywood screenwriting, social differences become linked to notions of “authenticity,” such that writers are only thought to only be able to create in particular ways, depending on their social backgrounds. On the whole, this arrangement tends to disadvantage women and non-whites, who are 96 frequently narrowly typecast as particular kinds writers only capable of writing more niche screenplays (Bielby & Bielby 1996). In business courses I observed on the talent industry, one agent remarked how African American writers sometimes got typecast as creators of “urban” pictures. Although I was not able to witness much of this race or gender-based discrimination first-hand, I heard of instances when a writer’s “symbolic” attributes played into him or her receiving work. At a screenwriting conference, one studio executive and panelist described how they had decided to reject working with a writer after meeting with him in person. “I know it sounds silly,” she explained, “but he was just sweating profusely throughout the meeting. And he was very excited and everything, but he always kept spitting when he talked!” These examples allude to part of the reason why meetings between writers and executives, producers, or agents became so important (as they were at Limelight). In these meetings, writers not only had to be able to convey their “passion” for their particular project, but also convey a sense of acceptable “authenticity” that was consistent with how they could present themselves as individuals. Furthermore, writers are also evaluated within the industry as people based on their previous experience working within the industry. Having previous credits serves as a kind of “credential” for those hoping to find more work as writers (Bielby & Bielby 1996; Baker & Faulkner 1991). Having previous experience allowed writers to more easily be represented by an agent, providing these writers a fundamental level of access over other writers who did not have credits. This presented a kind of Catch-22 situation for new writers seeking entrance into this “cultural field:” it was difficult to get writing credits without an agent, even while it was difficult to secure agency representation 97 without writing credits. One screenwriter I interviewed recalled his early experiences breaking into the industry, stressing the importance of having an agent. He had written a screenplay that he and his current agent were now “pushing” on the market. Yet, back when he first started, he had the same screenplay but did not have agency representation. He described an exchange in which he was able to meet an executive while waiting tables at a restaurant, and ultimately mail the executive his script for him to read. I was at home and I get this call and he’s like, ‘This is [so and so] from [name of studio]’ And I was like, “Ah.” And he was like, do you have an agent? And I was so dumb, I should’ve been like “oh, yeah” and just made some name up. But I was like, “No, I don’t.” And he said, “Ohhh… Well, that’s too bad. I can’t read your material then.” And then he just hung up. But like, my contact information was inside the envelope, so I know he opened it up and read it…. It’s sort of like, “oh, you don’t have an agent. You must not be a good writer.” Evaluation of writers based on their experience was partially mediated by agencies (Bielby & Bielby 1999). When screenplays were submitted to production companies or studios, they contained the writer’s resume and cover letter, both written by their agent on agency stationary. Agency representation thus allowed a screenwriter’s screenplay to be opened and read, but, as I discuss below, the quality and number of a writer’s credits could also heavily influence whether a particular screenplay was furthered in the moviemaking process. Intuitive Evaluations of Motion Picture Stories The third major way in which movie ideas were evaluated (my primary focus in this chapter) involved categorizing movie along several criteria, including their “commercial” appeal, their percieved “edginess,” and their “artisticness.” Implicit in 98 these categories are ideas of audience appeal—that is, notions of which films will appeal to which audiences and whether a movie will appeal to movie spectators in general. At Limelight and elsewhere, many of these categories were formalized in the “grids” portion of script coverage (see Table 1 above), even though I was told by several readers that this part of script coverage varied slightly depending on the company. Even so, these core categories represented key ideas that were mentioned in the examples of script coverage I read and in interview conversations I had with studio executives, script readers and analysts, and various screenwriters. They formed pieces of a shared language of audience appeal that all of these individuals were familiar with, in their various ways. As a way of highlighting how these descriptors of movie ideas constitute a shared language of evaluation in the motion picture industry, I mailed a brief survey listing hypothetical movie ideas to a sample of working studio executives. Following my ethnographic and interview fieldwork, executives were sent the survey, which contained eight hypothetical movie loglines to be rated on a scale of 1-5 (see Table 3). These movie ideas were intended to represent a range of different kinds of movies, based on what my interviewees said constituted “commercial,” “edgy,” and “artistic” movies. Survey respondents were also asked to rate the movie ideas in terms of their perceived “marketability,” a concept that was extremely closely related to the idea of “commercial- ness.” Although, I only was able to obtain a small number of responses (n=13), the responses are nevertheless suggestive (refer to the Methodological Appendix for additional information on this survey). 99 ________________________________________________________________________ Table 3: Survey of movie executives, loglines for hypothetical movie ideas ________________________________________________________________________________________________ ________________________________________________________________________________________________ Movie A: A lone alien crashes its ship on the lawn of the White House and has a secret that could save humanity from pending disaster at the hands of a rival species. A military scientist fights for the alien to be understood before it is too late. Movie B: One day, a large alien ship is seen heading toward earth, and, as it orbits the planet, no one knows why it has come or what the aliens want. As humanity debates what the alien arrival means, it eventually becomes apparent that aliens have been lying dormant in human bodies for a long time, and are now hatching to finally colonize the planet once and for all. During the conquest, a military scientist discovers that a rare microbe is an effective weapon against the aliens, and he undertakes a treacherous mission to deliver it to resistance fighters. Movie C: A serial killer with his own sense of justice kills those who escape the criminal justice system. The police detective who chases and eventually finds the killer is led in a powerful way to question the nature of good and evil. Movie D: A police detective chases a serial killer who has terrorized his city for months. As he unravels the crime spree, the detective realizes that the killer is closer than he initially thought: he is forced to bring his own brother to justice. Movie E: A group of young radicals succeed in kidnapping a corrupt President ruling over the United States in a near, dystopian future. In the government’s ensuing effort to rescue the President, a popular revolt is inspired and a new era of freedom is ushered in. Movie F: A terrorist group infiltrates the Secret Service and succeeds at kidnapping the President while he is on an overseas trip. While the terrorist group holds him for ransom and attempts to gain political concessions from the United States, a lone Secret Service agent on the right side of the law takes a daring chance at rescuing the President. Movie G: After mistakenly thinking that he has won the lottery, a young Mexican gardener and his coworkers decide to quit their jobs by playing a series of hilarious landscaping pranks on their shallow and self-absorbed employers in Beverly Hills. Upon realizing that his lottery ticket is for the wrong week, the gardeners are eventually able find common ground with their employers and come to a new understanding of the American Dream. Movie H: A young Mexican immigrant to the United States works as a maid at the estate of a wealthy family in Beverly Hills. Although she is initially caught up in what appears to be the family’s idyllic lifestyle, she gradually uncovers the family’s dark secret… which leads to a murder and sends the maid fleeing for home. As she crosses the border back into Mexico, the maid realizes that appearances are not always reality in the “land of opportunity.” ________________________________________________________________________________________________ ________________________________________________________________________________________________ Table 4 summarizes the responses these 13 executives gave to the movie loglines contained in the survey. What is most noteworthy is that responses to the loglines generally had fairly small standard deviations—especially when it came to executives’ 100 ratings of movies as “commercial” and “marketable” (two categories that were positively correlated). In other words, regardless of the movie idea, these 13 executives rated movies as “commercial” and “marketable” within a very close range. Standard deviations for ratings of movies as “edgy” and “artistic” were slightly higher, suggesting a bit less convergence of opinion when it came to evaluating movie ideas based on these criteria. ________________________________________________________________________ Table 4: Survey of movie executives, ratings of hypothetical movie ideas: Mean scores and standard deviations ________________________________________________________________________________________________ Movie: A B C D E F G H ________________________________________________________________________________________________ “Commercial” 3.09 3.30 2.70 2.50 2.56 3.44 2.22 2.11 (1.13) (0.82) (0.82) (0.84) (0.73) (0.88) (0.83) (0.78) “Marketable” 3.27 3.40 2.80 2.10 2.67 3.44 2.33 2.22 (1.01) (0.97) (0.92) (0.99) (0.71) (0.88) (0.87) (0.83) “Edgy” 2.00 2.40 3.10 2.30 2.89 2.67 1.89 2.66 (0.63) (0.84) (0.99) (1.25) (1.05) (1.00) (0.78) (1.12) “Artistic” 1.45 1.70 2.20 1.70 2.89 2.22 1.89 2.56 (0.69) (0.82) (1.23) (0.95) (1.27) (0.97) (1.17) (1.24) n=13 Respondents rated movie ideas on a scale from 1-5, which used the following values: 1= “not good”; 2= “so-so”; 3= “good”; 4= “very good”; 5= “excellent.” In addition, as the mean responses to each of the movie loglines show, no movie was rated especially highly on any measure. Even so, there were some clear differences between how intensely each movie logline was scored by the executives. Some movies (such as Movies A and F) were clearly rated as more “commercial” than others, for example. Similarly, different movies were rated as most “artistic” (Movies E and H). I 101 speculate on the meaning of these differences in more detail below. But, for now, it is worth noting how the executive survey responses showed how sentiments toward varying movie ideas both converged in their ratings of individual movies and varied between different movies. This suggests how Hollywood executives do not have their own idiosyncratic, random understandings of intuitive categories like “commercial,” “marketable,” “edgy,” and “artistic.” Rather, these ideas provide a shared language of evaluation for Hollywood executives. 4 Definitions of Intuitive Categories So, if intuitive categories do in fact provide a shared evaluative language within the Hollywood film industry, what exactly do these ideas mean? That is, how do they define movie ideas, giving these ideas attributes that help decide their fate as they move through the movie-making process? “Commercial” For my interviewees, the idea of a “commercial” movie meant a couple of things. In one sense, interviewees spoke of “commercial” as “whatever makes money.” In theory, then, any movie could really be “commercial.” However, “commercial” also referred to movies that shared common a common “look” and “feel” in terms of their visual elements, plots, and moral structures. 4 Along these lines, these intuitive evaluations may have particular meanings within the Hollywood film industry. When I gave this same survey to a relatively small group of undergraduates, their average ratings were sometimes very different than those of these executives (see Methodological Appendix for more on this comparison). 102 One part of what is meant by “commercial” movies involves what Justin Wyatt calls “high-concept” movies—movie ideas that have overt, transparent visual elements that can be summarized in a movie title and easily represented in a movie poster or its advertising (see Wyatt 1994). At the outset of his book, Wyatt quotes director Steven Spielberg, who famously said of movie ideas: “If a person can tell me the idea in 25 words or less, it’s going to make a pretty good movie. I like ideas, especially movie ideas, that you can hold in your hand.” Those I interviewed echoed Spielberg’s sentiment that “commercial” movies were heavily visual movies, ones that could encapsulate a great deal of the story in an image or picture. 5 One working studio executive put this as follows in an interview, pointing to a poster of Independence Day (1996) that hung on his wall (a movie he had worked on). Using this, he explained to me what “commercial” meant: An alien ship comes to this world to take it over. Will Smith is involved with it. It takes place on July 4th… It’s a movie where the White House is being blown up by this thing, this explosion that fills the screen. It’s like, “Ok, I get it. Tell me where I go to see the premiere!” One script reader I interviewed also spoke to the importance of visual elements in making “commercial” movies, particularly as they related to the comedy genre. She put it as follows: “In comedy, if you can write a comedy that is dependent on a lot of physical humor and a lot of gags—that is commercial. In that sense, Animal House is a hugely commercial comedy.” In a similar vein, one analyst wrote in a piece of coverage how a movie could be made too narrow in its appeal (and thus uncommercial) by a lack of 5 In this way, the notions of “commercial” and “marketable” are very close cousins. If “commercial” films are ones that are seen as reducible to a small number of transparent elements, then “marketability” refers to how easily those elements might be represented in advertising. Thus, it is not surprising that the movie ideas summarized in Table 3 all received very close scores when it came to how “commericial” and how “marketable” they were. 103 reduction to a simple, transparent visual element. She described the script (a science fiction script that she ultimately “passed” on) as follows: The material has a loyal (fiercely loyal) audience but it is a niche audience—MTV viewers who are also fans of both science fiction and animation. This script seems targeted to that niche demographic, without any real thought to broadening its appeal…. With this sort of ensemble piece, it’s often hard to come up with a coherent logline, one sentence that can embrace everything that happens. In a related vein, “commercial” movie ideas had relatively simple plots, which included, among other things, a “fast set-up.” One script analyst I spoke with explained this in an interview, using the example of The Fugitive (1993) as a kind of archetypical “commercial” movie. If you’re speaking in general about “commercial,” The Fugitive is a highly commercial movie. The Fugitive with Harrison Ford. It has a fast set-up, the story is told through images and actions. By the end of Act I, basically you have a very strong and very sympathetic hero. He’s victimized for awhile, for awhile he is only on the run. At some point, because of what happens in the premise, he starts acting on his own behalf to extract himself from what’s going on around him, and he solves the conflict himself. The Fugitive thus had the “look” of a commercial movie, and was paced like a “commercial” movie because it provided a very quick lead-in. The above description also alludes to how “commercial” movies are defined in part by their moral structure and how this is actualized through a movie’s characters. In this vein, “commercial” movies have to involve an individual, proactive hero. One reader attested to how this was important to her as she covered scripts. She recalled a script she had just covered the night before. Last night on one of the scripts I was covering, we had a hero and she was a rape victim. It really ended up being a bad made-for-cable movie. And it featured a rape victim whose rapist was brought to trial, her father 104 suffers a heart attack from the stress, the guy goes to jail after pleading to a lesser charge, and he is released four years early and comes after her. And basically it piled injury upon injury upon injury. And the hero was sympathetic, but she was not proactive. She was a victim—victim with a capital “V.” And very, very late in the story, she was backed into a corner and only then did she start fighting back. And that was one of my major complaints—that the hero was not driving the story. She went on to explain how this principal applied to the villain as well, the force of “evil” in a movie. …. And that’s another element of a commercial film. You not only have a well-defined, proactive hero but also a well-defined proactive villain as well. Even if the villain is a force of nature, a plague of locusts, Frankenstein’s monster, a corrupt politician. Thus, “commercial” movie ideas involve clear, unambiguous, proactive heroes and villains who exist in dynamic tension with one another, with the hero always eventually obtaining a full and complete victory in the end. In this way, readers’ and analysts’ ideas of “commercial” echo ideas set forth in Christopher Vogler’s influential manual for screenwriters, The Writer’s Journey: Mythic Structure for Writers (1993). Both widely read by screenwriters (it was repeatedly mentioned in interviews and screenwriting conferences I attended) and notably critiqued by several academics, Vogler argues that many commercially successful Hollywood movies have tapped into deeply-held “myths” involving heroes and villains that are universally held across human cultures. Accordingly, heroes and villains represent universal types, with stories that tap into this mythic sensibility being driven by a lone, individualistic hero struggling against adversity and ultimately being made whole at the movie’s end. I also detected hints of this ideology about why movies resonate through the executive survey I conducted (see above). The two most “commercial” loglines 105 (Movies B and F) involved this sort of uncomplicated, “mythic” storyline—a morally good individual struggling against adversity to ultimately defeat a clear “villain” (in this case, a terrorist group and an alien army from outer space). Indeed, it is simultaneously interesting that the 13 executives I surveyed rated Movie G (the one about the Mexican maid fleeing for home after discovering a “dark side” of the American Dream in Beverly Hills) as the least commercial idea. Here, there was no ultimate triumph of an individual hero over adversity, but only an escape where the hero fell short of being made whole. In addition, the “villain” perhaps seemed less clear and less universal, tingeing the movie logline with more moral complexity and ambiguity. A final defining feature of “commercial” movies dealt with their perceived “accessibility.” In this light, “commercial” movies could not be too “intellectual” or too “serious.” In this sense, the notion of accessibility contained an important assumption about mass audiences: that they would only like fare that was billed as unintellectual and “light.” One script reader explained “accessibility” in the following way during an interview: Sort of an unspoken criterion is “accessibility.” It could be a great premise, but not accessible as a movie… Accessibility means, how accessible is it to an audience? Is it going to reach an audience? Is it too esoteric? It may be a great idea, but is it going to really translate into entertainment? So “accessibility” is sort of “entertainment value.” Given this understanding of “entertainment,” various factors could cast an idea as too “intellectual” to have broad appeal or too “serious” to be enjoyable to the masses. For instance, it was more difficult to see certain genres and settings as “accessible.” One reader told me how this was especially true of “period dramas.” 106 Period dramas are not considered “commercial.” Most big companies like Warner Bros., they won’t touch something like that unless it has some very strong action element to it and/or it has some very strong directing and acting talent that’s signed onto it. In addition to the genre and setting potentially making movie ideas “inaccessible,” the dialogue and overall tone of a movie idea could help construct it as lacking “commercial” appeal. One script analyst who I interviewed provided me her coverage for one such movie idea. In her coverage report, she described the movie as “serious and serious about its issues” and as “a script that has something to say.” Eventually, the script went on to be made into a critically acclaimed film with only moderate commercial success. And, not coincidentally, it was never marketed as a “commercial” would have been. In fact, this trajectory for the movie seemed almost preordained in the analyst’s coverage of the movie. She characterized it as follows: The dialogue is very good, but there are places where it seems almost “too good.” That is, the speeches go on just a little too long and they’re just a little too “intellectual.” It’s not that the language is pretentious—it isn’t— but the men at the center of the story are wordsmiths—language is their life and they love playing with words, sometimes because they like the sounds of the words as much as the meanings of the words. Thus, because the film had a little too much to say and said it with a little bit too much flourish, the analyst deemed it as “inaccessible.” When I interviewed her, this same analyst relayed a similar set of explanations in telling me why she thought some commercially successful books did not translate to the screen. She explained: It may be accessible as a book, but not as a movie. You just don’t want to watch it…. A perfect example is Angela’s Ashes. How do you say “no” to something that’s sold so many copies and so widely read? It just doesn’t translate into a movie. Too dreary, too bleak, too dark. 107 Thus, the idea of “accessibility” ultimately had to do with perceptions of breadth of audience appeal. Assumptions about what made movie ideas accessible held that any number of factors might be introduced that would limit the size of the audience. If a movie idea was defined as too serious, too intellectual, or too eloquent, it tended not to be viewed as “commercial.” In fact, however, it is worth remembering that how movies have been constructed as “accessible” and “commercial” has changed substantially throughout the history of American movies. As discussed in Chapter 2, movies were at one time talked about as having commercial potential precisely because they were seen as having an intellectual, “high-brow” appeal (Baumann 2007). In addition, there are notable present-day examples where “uncommercial” movies nevertheless become commercial successes. As a language of evaluation, then, ideas about commercial appeal form a key argument about why movies “work” and why movies should be made, though the specifics of this argument have proven to shift over time. “Artistic” The notion of an “artistic” movie largely stands in opposition to the notion of a “commercial” movie. Elements of movie ideas that are thought to limit the “commercial” appeal of movies are some of the same features that help construct them as “artistic.” Where “commercial” movies are seen as having an essence that can be reduced to an image or a short logline, “artistic” movies escape this reduction and have stories that are seen as more “complex.” Additionally, while “commercial” movies are seen as emphasizing physical action, “artistic” movies do not. They are instead conceived of as 108 focusing on more subtle, non-physical action, offering richer characterization. Furthermore, more “high-brow” cinematic devices—including formal experimentation or dialogue that is in non-contemporary/non-standard American English—send signals that a movie is meant to be “artistic.” Given their emphasis on more subtle action, “artistic” movies may be easier to accomplish in particular genres (at least as they are currently understood). Many films falling in the “independent” or historical drama genres, for example, are frequently described as “artistic.” The pole between “commercial” and “artistic” was thought to map on top of a pole between “mass audience” and “niche audience.” In addition, there were implied demographic realities that surrounded both the notions of “commercial” and “artistic,” realities that confounded this simplistic dichotomy between broad “commercial” audiences and narrow “artistic” audiences. Specifically, the implication of “artistic” movies was often that they were more mature than “commercial” movies. Herein, the difference between “commercial” and “artistic” was not so much a difference between “mass” and “niche” as it was between “young” and “old.” “Edgy” The notion of “edginess” transcends the divide between commercial and artistic fare. Sometimes, a film can be simultaneously “edgy” and “commercial” while another movie might be “edgy” and “artistic.” “Edginess” alludes to how a movie might have a genuinely raw feel. When a movie is described as “edgy,” it also may be described as having more in-your-face, even distressing, imagery. “Edgy” films are meant to portray a gritty reality laden with intense, real emotion, and perhaps even forceful social 109 commentary. Some of the screenwriters and agents I observed and interviewed spoke of how they personally preferred “edgier” fare, so much so that they wished that the contemporary movie audience sought “edginess” as much as they themselves did. Along these lines, they spoke of Hollywood movies from the 1970’s as representing a creative apex from which American filmmaking has since declined. One team of writers spoke to this at a panel on selling scripts at a screenwriters I attended. The pair had recently sold a script that had become a hit comedy. One of them told their audience of fellow screenwriters, We love ambiguous films. In the 70’s, they made hugely ambiguous movies that surprised and shocked. But, you know, we write mainstream Hollywood comedies here. There’s only one ending. You just try to come up with something you don’t hate. Even given the oft-stated personal preferences of moviemakers for “edgy” movies, varying degrees of “edginess” could make a movie more or less “commercial” (that is, more or less suitable for a broad audience). At one extreme, movies that were seen as “too edgy” were also seen as not appropriate for mass audiences. A movie may simply be too jarring—either visually or thematically—to remain entertaining for a broad audience. If this happened, a movie idea might be relegated to arthouse cinema or cast as a movie for some other “specific” audience (e.g., horror). To avoid this extreme, then, the idea was that movies should appropriately wear down their “hard edges.” Ideally, “commercial” movies that strove to be “edgy” should above all be “authentic.” They should maintain a real, plausibly gritty feel, but should not cross a line of disturbing the common moral sensibilities of mass audiences. Thus, commercially edgy movies might depict harsh realities of the contemporary world, but should still, at some fundamental 110 level, allow uncontroversial heroes to triumph over uncontroversial villains. In the end, there should still exist a triumphant struggle against adversity, with good clearly winning over evil. 6 At the other end of this spectrum, movies might not have enough “edge” to be commercial. Examples of this would include movie ideas that are intended to be “edgy” but do not succeed in plausibly executing this “edgy” feel. When this happens, a movie idea comes off as cliché, “corny,” “hokey,” “over the top,” or “cheesy.” Such movies are movies that may have been intended as “serious” movies, but may end up as jokes whose clips circulate on YouTube. As one agent involved in the script market said in a class I observed on the movie business, “you want to read something that is the answer to that.” Making Intuitive Knowledge at Limelight Entertainment When I think of the work of many executives, producers, and talent agents in the business of moviemaking, I sometimes think of a collage that artist and cultural critic Barbara Kruger made. Featuring short, caustic statements placed on top of photos from old magazines, Kruger’s collages are influenced by advertising and serve as blunt, militant, if sometimes cynical commentaries on contemporary culture. The collage I remember here featured a black and white close-up of two male hands clasped in a firm handshake, with the words (in bold letters), “You make history when you do business” layered over top of the image. With this, Kruger seems to be making a jab at the sense of masculine grandiosity she saw in the business culture of corporate America. 6 In this light, it is interesting to note the results of the executive survey mentioned above. Of all the movies that the 13 studio executives rated, the “edgiest” was Movie C, a hypothetical film about a serial killer with “his own sense of justice” who murders criminals who have escaped the law. The conclusion of the logline alludes to how the cop who chases this serial killer is forced to question the “nature of good and evil,” which probably contributed to this movie’s relative rating as “edgy.” 111 While bracketing the overall tone of Kruger’s collage, over the course of my fieldwork I did come to see the superimportant, almost mythical way in which moviemakers talked about the business of making movies. This was especially true of talent representatives, who were among those I was able to observe most closely. Their theories of how the business worked were inhabited by a variety of mythical success narratives. They had a shared memory involving a common cast of characters who “made it” in the business, people and agencies who rose and fell, the biographies of film industry “legends” that sometimes worked as morality tales for others in the business. The talent representation business was built by men such as Lew Wasserman and Michael Ovitz, men who were at one time seen as powerful kings ruling Hollywood but who could fall very far very fast. There were now powerful players who rose from obscurity based on the strength of their business savvy. The specific circumstances of why people rise and fall in this world vary, but an ultimately decisive factor in these success narratives is how well individuals understand the “rules” of Hollywood—the shared interpersonal culture within which deals are struck and decisions about movies are made. This interpersonal culture, with all its attendant rules, forms an important backdrop against which intuitive knowledge about audiences is constructed. In the case of the Hollywood talent industry, what most defines its interspersonal culture is the character of this industry as a hierarchical organizational “field” in which there is a limited, core set of firms and individuals that are constructed as being most legitimate, and a more peripheral set of individuals and firms who have less legitimacy. In the talent industry, legitimacy is secured through the positive perceptions of other players—one’s 112 peers within a dense interpersonal network. Given the realities of living and working in this world, individuals and firms constantly attempt to affirm their legitimacy by performing what I have called “reputation work, ”or the intentional activities that participants do in order to create the perception that they are legitimate, according to institutionalized expectations operating at the industry level. The concept of reputation work highlights how “impression management” is governed by institutionalized rules that are constructed at the level of the organizational field. Reputation work occurs as talent agents and managers engage in strategic self-presentation in order to conform to industry- level expectations for appearing legitimate. This organizational context sets up a culture of dealmaking in which intuitive explanations for why movies work become convenient and expedient. As I discuss below, this intuitive knowledge—based on a principal of inspired judgment—is readily used by individuals seeking to advance and protect their larger reputations in the industry. On the other hand, technical knowledge about audiences (discussed in the following chapter), is much less resonant in the world of talent agents and managers (even though it could conceivably be extensively used to make judgments). This is because it is held as more “objective” and less malleable, it is less readily spun (and perhaps less understood), and much less useful for building relationships in Hollywood’s dealmaking culture. Importantly, an intuitively based argument need not be “believed” by those speaking it, but must only represent a plausible frame for interpreting movie ideas. In this way, discussions about movie ideas based on moviemakers’ intuitive knowledge may amount to a kind of “communication distortion” in which legitimacy trumps all else—situations 113 in which legitimate industry players make legitimate arguments about movies, but where the ultimate conclusions about movie ideas may be inconsistent. In the sections that follow, I discuss how reputational games and intuitive knowledge worked in concert to shape how decisions about movies were made during my observations of the Hollywood talent industry. This work was based primarily on my observations of Limelight Entertainment (a talent management company also involved in producing movies), but was also supplemented by a business school course I observed on the “talent representation” business. First, I describe how the realities of “reputation work” powerfully shape the contexts in which deals about movies are made. Next, I show how the intuitive language for evaluation discussed above worked as a legitimate discourse for evaluating movies at Limelight. Finally, I discuss of how these two factors interact, helping to reproduce intuitive knowledge in this segment of the Hollywood movie industry. “Reputation Work” at Limelight A typical day at Limelight Entertainment begins by 8:00 a.m., when all of the assistants have arrived. Drowsy and efficiently sipping their coffee, they page through their email accounts and orient themselves for the day. Copiers and printers start humming to life as scripts, contracts, and client schedules flow in via PDF file and fax. And the multiple phones around the office start ringing, and, on some days, Limelight’s assistants do not stop answering them until 10:00 p.m. By 9:00 a.m., many of the managers start arriving, frequently in the midst of a cell phone conversation. Their work days have already begun, and they usually never 114 completely end as they answer emails and talk to clients on their phones well into the evening. Limelight’s assistants organize the frenetic schedules that fill their managers’ time and blur boundaries between home and work. Tom Young’s schedule on one typical day, for example, included workout time at 7:00 a.m., lunch with a client, a meeting with other managers at Limelight, his child’s soccer game, and a movie premiere in the evening. When they enter the office, few managers acknowledge the front receptionist or bother to formally greet their assistants. Instead, after they swipe their ID card in front of the electronic security scanner and enter the building, most managers just shout across the office to their assistants (“What do we got, Sandra?”). Between meetings and personal obligations, managers’ days are a stream of phone calls. The managers are equipped with telephone headsets, which many use as they walk around the office in a focused trance of negotiating deals and talking with clients. Between phone calls or before a set of phone calls, managers would sometimes shout out a list of calls for their assistants to dial as though they were mapping out a chain of possible events in their head. The assistants would dial the calls in rapid succession, often resulting in an involved game of “phone tag” with the other party’s assistant. On several occasions, the pace of a manager’s phone conversations became so frenetic that a manager would get someone on the other line but forget why he or she had ordered the call dialed (at which point the manager’s assistant would have to remind him or her). Thus, a manager’s day largely involved navigating a network of contacts in order to gather pieces of information about prospective film and television projects, and to try and direct these projects to move in particular directions—ones that would benefit their clients. 115 Forms of Reputation Work in the Hollywood Talent Industry In the course of navigating social networks and interacting with clients and business associates, Hollywood talent representatives consistently perform several kinds of reputation work. Managers and agents engage in these ongoing efforts in their attempts to foster the impression that they are competent, as judged by industry-wide institutions dictating how Hollywood talent representatives should look and act. Four common domains within which talent representatives perform “impression management” are explored in detail below and summarized in Table 5. These domains refer to the institutionalized expectations that talent representatives strive to ultimately meet in their front stage performances for industry players that they wish to impress. The reputation work required to meet these expectations involves the ways in which agents and managers work to maintain reputable physical environments, how they collectively construct the emotional tone in which their business is conducted, how they strive to follow Hollywood’s conventions of giftgiving, and the ongoing work that they do on their selves. When reputation work is performed adeptly (i.e., when it fulfills institutionalized expectations for how talent representatives should look and act), agents and managers see this as improving their reputation within Hollywood’s social networks. However, when reputation work is done poorly, talent representatives’ front stage performances fail to meet institutionalized, industry-wide standards and thereby expose agents and managers to reputational risks. 116 ______________________________________________________________________________________ Table 5: Institutions of reputation in the Hollywood talent industry ______________________________________________________________________________________ Domain of “impression Institutionalized expectation for Disreputable form “management” front stage performance ______________________________________________________________________________________ Physical environment located in/near Beverly Hills any other geographic location “sleek,” ordered, and disordered, cluttered “feng shui” Emotional environment casual; hip too formal or stiff; not hip Giftgiving creative and unique; chinsy or mass produced; appearing to be impersonal personalized Selfhood confident; socially adept socially awkward Men aggressive; confident swagger; submissive ownership of space Women warm interactional style; cold confident, but “nice” ______________________________________________________________________________________ Reputation Work and Making the Physical Environments of Talent Firms Early on in the course of my research, I had managed to interview a talent agent at Creative Artists Agency, which was at that time (and still is) the largest agency in Hollywood. This trip was one of my first times inside a talent firm, and it was my first opportunity to see one of these companies from the inside. The lobby of the CAA building was impressive. The vastness of its space, its glossy stone floors, its sensation of orderedness, and its carefully arranged collection of art made CAA’s lobby reminiscent of some museums. The lobby’s most famous work—widely known among those working close to the representation business—is Roy Lichtenstein’s 27 by 18 foot 117 “Bauhaus Stairway,” which, because of its immense size, was painted inside the lobby itself. The aesthetic of the CAA building did not happen by accident. Like the physical spaces of other major Hollywood talent firms, it was carefully planned. When it was initially constructed, for example, Michael Ovitz (who was then head of CAA) remarked to a reporter that the building was “built to be a piece of art itself” (Johnson 2006:1). The physical setting of a Hollywood talent firm constitutes a front stage performance—perhaps the most immediate of all front stage performances that talent representatives put forth to their clients, other talent representatives, and various other players in the Hollywood film business. Yet, considerable back stage reputation work goes into producing this reputable front stage environment. This work on the Hollywood talent firm’s physical space takes place on two levels. The first is the geographic location of the firm itself. In order to operate a reputable firm and be a player in the world of deal making, a firm has to be located in the right area of metropolitan Los Angeles. As Tom Young put it during an interview in which he talked about the significance of Limelight’s location, “I don’t think that you can replace the value of having a nice place like this, being anchored in kind of a nice building at this address in Beverly Hills.” Tom and Limelight’s three other partners had carefully selected the location of their office when they launched the company. I was allowed to see old files which showed that Limelight’s partners had considered several different locales for their new company, photographing prospective sites in advance before leasing a space. In the end, after looking at over two dozen potential business sites, Limelight’s partners decided on a location that was in close proximity to other major talent firms in Beverly Hills. This geographic presence afforded Limelight the status of a reputable industry player, and provided legitimate 118 access to the networks in which deal making takes place. Important in this regard are prominent restaurants located near Limelight, which the Zagat Survey of Los Angeles dining rates among the city’s top “power scenes” (see Pettera 2006). Thus, Limelight’s managers perceived the company’s physical location as providing it with necessary entree. According to Tom Young, it helped distinguish the company from what he called “mom and pop companies” out in Burbank or the San Fernando Valley. Figure 10: Sculpture near the entrance of the International Creative Management building in Beverly Hills. The second level at which talent representatives work to meet institutionalized expectations for a reputable physical environment is by designing and maintaining an appropriate interior aesthetic. Hollywood talent firms can vary somewhat in their specific designs. During the course of my class at Pacific Talent Agency, students were sent out to do projects at several prominent companies. Upon reporting back to the class, 119 the course instructor always asked students to talk about “what the place looked like.” The variation in the interior designs of talent firms becomes part of their attempts to forge unique corporate images. CAA’s art museum-like feel is one possibility. Others, such as PTA, have entirely different color schemes and use different stones in their floors and walls. Limelight, as another example, had what I heard managers refer to as an “industrial” aesthetic. This included bare concrete floors, exposed air vents, and a concrete ceiling high above the work space. The flattened remnants of metal beams poked through the floors, which I heard a manager tell a screenwriter client was done intentionally. Thus, Limelight’s interior was purposely designed (with the help of a consultant) to look fashionably worn, like a pair of expensive designer jeans that have been purposely frayed. Although they differ in their particular designs, a common denominator among the aesthetic of Hollywood talent firms is their sleekness. Offices cannot be cluttered, as those of many university professors are. There can be no loose papers and no loose cords. There can be little visible evidence of physical records, despite a vast quantity of them. This results in talent firms having a certain minimalist orderedness in their appearance. CAA was even designed and dedicated with the help of a feng shui consultant from China. Limelight’s managers appeared to also be influenced by similar ideas relating to the creation of the physical environment in which they worked. Aside from the orderedness of Limelight’s work space, Eastern symbols were displayed in certain parts of the office (including a fairly large statue of Buddha in a conference room). Limelight’s managers also hired a “building consultant” to periodically come to the 120 company and perform various rituals meant to improve the space’s “energy” (this will be discussed in more detail below). On the other hand, the offices of talent firms cannot appear to be merely functional either, like a bland collection of cubicles acquired at least cost. There must be some intention to the design, some creative vision driving it. The aesthetic of talent offices must be coordinated to look “industrial,” or like an art museum, or in accordance with some other unifying theme. Various decorative touches added to the respective aesthetics of the firms that I saw. One manager at Limelight, for instance, had a collection of small antique movie cameras arranged on a table in the corner of his office. Elsewhere in Limelight, there were original black and white photographs bought from an area artist, as well as vintage movie posters framed in the bathrooms. Managers at Limelight and agents in other firms would often hang in their offices posters of shows or movies they were involved with, decorations that undoubtedly were designed to favorably impress clients and other visitors. All of this effort came back to making a firm and its talent representatives appear reputable, as legitimate players among competitors. As Tom Young put it when I interviewed him, For us, we want to make sure we have the kind of clout and relationships and reach in the business that we are in play with important actors, or writers, or directors who take meetings. And there are various competitors around town who have their own creative aesthetic environments, and I think it’s key for them and, you know, key for us. The wide open nature of the workspace out there, the high ceilings, I think it’s a really good creative space for actors, potential clients—whomever. They seem to really respond to it. The course instructor in my class at PTA made a similar point when he noted (with a bit of hyperbole) that prospective clients were sometimes so impressed with the company’s building that they “signed them right in the lobby.” At Limelight, as well, prospective 121 clients and other business associates sometimes commented how they liked the office’s “industrial” look. In this way, part of a talent firm attaining “clout and relationships and reach,” as Tom Young put it, involved the extent to which the firm’s front stage physical aesthetics could be used to favorably impress others in the industry. At Limelight, once the initial conceptualization of the design was in place, a constant set of back stage efforts were undertaken to maintain the company’s aesthetic. Doing this back stage reputation work was a way for Limelight’s managers to deal with reputational risks. If having an impressive workspace contributed to the firm’s favorable reputation in Hollywood, then the possibility of the office becoming disordered was simultaneously the possibility of Limelight’s reputation being lessened. A couple of Limelight’s partners therefore took a particular interest in maintaining the sleekness of the office, directing the company’s assistants to perform much of this work. Dan, a Limelight partner who Tom Young referred to as “our resident queer eye” (a reference to the TV show Queer Eye for the Straight Guy), led much of this effort, tweaking the layout of furniture every couple of months, organizing “cleaning parties” for the assistants (late in the evening, after the phones stopped ringing), and routinely regulating the possibility of clutter piling up. At times, this conflicted with the practical demands of assistants’ work. For example, a collection of wooden file cabinets formed a large rectangular table near the front window of Limelight’s office, which faced out onto a major avenue in Beverly Hills. When some of the assistants began using the table to assemble the many documents that flowed through the company, Dan told them, “Guys, this is not a work space.” When Mary, one of the assistants, asked why not, he told her (as though it should be obvious), “Because people driving by can look in and see the 122 stapler and hole puncher.” This consciousness of how the building looked to those driving by even extended to the public sidewalk outside Limelight’s office. When anyone who appeared to the managers to be undesirable would linger on the sidewalk in front of the office, including a group of boys skateboarding or the rare homeless person, one of the partners always directed one of the assistants to go outside and tell them to move away. Much of the back stage reputation work that talent representatives do to maintain their offices would appear petty in many other workplaces. An example of this was how food and drinks are prepared for visitors to talent firms. One agent who spoke to the class at PTA was amazed at how much time she spent arranging fruit trays when she was an assistant. “Here I am with an MBA and a law degree, and I’m checking each grape to make sure it doesn’t have any spots!” Similarly, it is a knowing joke of sorts that drinks offered to those who visit talent firms have to be eclectic. Firms maintain a variety of beverages in their offices, and the coffee, soda, and bottled water all should consist of more expensive brands. One member from a duo of screenwriters visiting Limelight played this joke on me when I brought him a bottle of H2O brand bottled water. When I handed him the bottle, he said, “No, no. I only drink Fiji water,” playing on a stereotype of the high-maintenance “Hollywood type.” None of this presentation of the physical space really comes down to a question of the bottled water, the fruit, or the appearance of a firm’s office per se. Rather, it is a matter of what all of these elements of a firm’s physical space represent about the firm. Talent representatives work to maintain an ordered, impeccable physical environment because that is what reputable firms working in Hollywood do. Worrying about this—that is, working at producing the desired “front” 123 for clients and business associates—is therefore part of maintaining the company’s reputable status in the entertainment industry. Reputation work and the emotional tone of talent firms A number of studies of cultural dynamics in group settings have shown how people have collective understandings of how group interaction should take place, and understand broader rules for engagement in particular kinds of groups (Magnuson 2005; Eliasoph & Lichterman 2003; Eliasoph 1998). Members of religious or civic groups, for example, collectively understand “what kind of group this is” that they belong to, and adhere to institutionalized “group styles” that set parameters for participation and conduct within that particular group (see Lichterman 2005; Eliasoph & Lichterman 2003; Becker 1999). Insights from these studies can be extended to an analysis of business settings, which can have institutionalized emotional tones that are shared across different firms (see Hochschild 1983). Some workplaces are expected to feel formal. At other workplaces, group members are careful to be polite and take great pains not to offend others. In Hollywood talent firms, institutionalized rules exists for how talent representatives and their clients should collectively work with each other. These rules set the terms for the emotional tone of work at talent firms and interactions between talent representatives, their assistants, and their clients. According to these rules, work at talent firms should feel casual, hip, and even unconventional, and talent representatives and their assistants collaborate as interactional teams to produce this overall mood within talent firms (see Goffman 1959). I noticed hints of these conventions early on in my 124 fieldwork at Limelight. Anticipating a more “business-like” dress code, I had invested some money in a new shirt and tie prior to my first meeting at Limelight. At the end of this meeting (in which I explained my research project), one of the assistants told me nicely what I had already figured out by looking around—that I was overdressed. “Yeah, we’re casual here,” she said. Modal attire for Limelight’s managers and their assistants was a pair of jeans and a t-shirt or casual shirt or blouse. But she seemed to be referring to more than just the dress code. The tone of my initial meeting at Limelight seemed very relaxed. I had prepared for a more formal exchange (one resembling a job interview), but the entire meeting was very conversational and people at the meeting seemed very friendly. They did not want to sit upright in their chairs and ask pointed questions, but instead wanted to curl up on their couch and chat. The dress code at Limelight was perhaps meant to serve as an appropriate backdrop for interactions between managers, assistants, and clients. All of these individuals seemed to constantly be at “work,” but it was as though they did not want to bother with unnecessary formalities. It was more important that everyone seemed comfortable and loose while they were at Limelight. As one manager put it when describing the company to a prospective client, Limelight offered a “homey” environment. Importantly, this casual environment was part of the experience that Limelight attempted to project onto the front stage in order to help set the tone of interaction with clients. Actors and screenwriters who came to Limelight’s offices to conduct business had free reign of the space, and would go back to the kitchen to grab some food or hang out with the assistants and chat. Both clients and managers often 125 brought their children to the office for several hours, and the assistants (or I) would spend time watching and entertaining them. The emotional tone at Limelight was often playful. On my first day, I noticed there was a large putting green at the back of the office. When I asked Tom’s assistant Mary about it, she said, “Oh yeah. The guys are obsessed with golf.” As I continued working at Limelight, I saw what she meant. The green was put to fairly constant use, along with a large net located in a meeting room which was used for practicing full golf swings. Male managers and occasionally their assistants would putt or chip golf balls while they talked on their headsets, conferred with other managers, or met with clients. Sometimes, meetings with male clients were entirely held in and around the putting green, which created an almost surreal experience of seeing celebrities from some of my favorite movies just putting golf balls. Other varieties of balls were thrown and bounced as well: footballs tossed between managers, small rubber bouncing balls off walls and into cubicles (sometimes targeted at assistants), basketballs dribbled, and hackey-sacs kicked—especially when kids were in the office. Tom Young was particularly fond of hockey, and he would occasionally bring out his hockey stick and swat balls around with this as well. The casualness of Limelight’s atmosphere also meant that things happened at Limelight that would have seemed inappropriate in other workplaces. Swearing and many off-color comments were not only seen as not offensive, but they were expected. Managers and their assistants swore prolifically—and not only when they were angry or excited, but in the course of normal conversation. Managers and assistants took breaks to look at various websites that might have seemed distasteful at other workplaces, 126 including one, pervertedjustice.com, which featured profiles of child molesters who stalk the Internet, or a website showcasing “Hollywood’s hottest assistants.” Sometimes, managers and clients made comments that had strong sexual overtones, as when a male client and a female assistant were joking about the size of a prospective actress’s “tits.” My fieldnotes record a number of examples of this. In one instance, Tara (an assistant), was the subject of a conversation that would have been considered “sexual harassment” in most workplaces: Tara found a screwdriver as everyone kept talking, and went to unscrew the attachment on the toy for Dan’s son. Troy Gordon [a screenwriting client] was like, “Now, you see this girl is from the Midwest with the way she uses that screwdriver.” He remarked about how she knew what she was doing, how she could use this tool with ease. He said something about this coming from her having grown up in cornfields. Tara was like, “You may not believe this, but where I grew up was literally out in the middle of a cornfield. I used to play hide-and-seek in the cornfield.” She was from Michigan, I remembered. Dan [a Limelight partner], was standing there holding his young son, and quipped to Troy Gordon, “I bet she did more than play hide and seek in the cornfield.” The sexual innuendo was clear here and they laughed, although Tara just seemed to ignore it and the conversation just sort of flowed on. Thus, although managers at Limelight worked constantly, the backdrop for this work was casual, sometimes with sexist overtones. This verbal misogyny was not only part of Limelight’s casual setting, but was simultaneously a performance of a gender hierarchy that had negative consequences for women at the company (this is discussed more below). Limelight’s employees were often joking or playing diversionary games around the office, and this playful environment blended into the fabric of their business activities. This, however, did not mean that Limelight’s atmosphere was emotionally calm. On the contrary, the mood around the office was frequently elevated as managers 127 yelled at their assistants or yelled at those they spoke with on the phone. But the work atmosphere was casual enough that these outbursts were expected (even mandatory). And it was also casual enough that relationships between managers and their assistants often appeared non-hierarchical. For instance, sometimes directives from managers would be met with smart backtalk or mockery from assistants. As one example, when Tom Young complained to Mary about the way she was handling his calls in a way she thought was petty, another assistant asked him in a smart tone, “why don’t you just fire her, Tom?” This “casual” group dynamic at Limelight could also be seen as representing another kind of interactional convention. Paradoxical though it was, managers at Limelight practiced a kind of conventional unconventionality. It was noteworthy that the bouncing balls, the off-color comments, and the sometimes outlandish behavior of managers all were such integral parts of Limelight’s work environment—so much so that they appeared mandatory. In this way, there was an intentionality to managers’ casualness. Thus, the emotional tone at Limelight was often one of extremes, sometimes carried out using joking and irony. Following Goffman’s understanding of interaction, these extremes could appear intense but simultaneously not be deeply felt by managers producing this group context. All of the antics, all of the ways in which talent representatives appeared to not pay much heed to the rules of typical office environment, was actually part of their work. A second noticeable element of the atmosphere at talent firms is what might generally be called their “hipness.” There was a definite “lookism” among those who worked at Limelight, and they practiced this on themselves and made it a prerequisite for 128 those seeking entrance to the company. All of the assistants at Limelight were in their twenties, attractive, and stylish dressers. Although they made relatively little as assistants, more expensive and current clothing brands were the norm. Most clothes from Target or Walmart would have seemed out of place, as would have many clothes more than a couple years old. This was part of a “look” that aspiring managers at the company had: not plain or frumpy, but not too quirky either. A couple months after beginning at Limelight, Tina (one of the company’s female managers) was hiring a new assistant, and the company brought in approximately twenty interviewees for the position. Most of the interviewees mirrored the physical profile of the company’s existing assistants. A couple interviewees, however, clearly did not seem to “fit” when they showed up in person. One was a woman who looked as though she was in her late 40’s and was dressed in a somewhat dated business suit. I noticed that her interview was over very quickly. When she left, Tina came out of her office, shook her head, smiled, and sighed, “Oh boy…” Sandra, the assistant standing next to me and who was responsible for lining up many of the interviews, apologized to Tina with a knowing chuckle. In this way, selecting employees who fit with the Limelight company look was a way of setting the stage for company’s atmosphere as a whole. How those working at Limelight looked was one way in which they produced a “hip” atmosphere. But this atmosphere was produced in a myriad other subtle ways as well. They included having a wide repertoire of styles used to greet and talk with business associates and clients. In greetings, these including the traditional handshake, the warmer kiss on the cheek, and a stilted version of a hand clasp loosely borrowed from hip-hop culture. It included a similar range of dialects as well, from traditional business- 129 speak recognizable to “the corporate world” to an inauthentic, at times ironic attempt at “urban” slang (“What is up, dude?”). Many of the managers often spoke in this way to their assistants and while talking to their clients. It even extended to the call hold settings. When one was put on hold while calling Limelight, the music was not the soft elevator music one gets when they call an airline reservation system, but a collection of indie rock hits that one of the assistants compiled. Which interactional repertoire was used depended on who a manager was interacting with. For instance, one polished looking older man wearing a business suit and talking with a French accent came to visit Limelight’s offices and needed to wait for an extended amount of time in the company’s office. Tom Young greeted the man with a traditional handshake, and in one of his more subdued demeanors, offered the man something to drink and politely welcomed him to ask for anything he needed (“You’re welcome to use my office to make phone calls, if you like.”). This more formal mode of interacting with “business types” was rare, however. Few people who visited Limelight dressed and presented themselves like the man with the French accent and business suit. Most visitors to Limelight were clients, prospective clients, or creative collaborators with clients who brought out a much more casual interactional style from managers. Indicating another way in which the interactional styles of managers could vary, the rapport that Limelight managers had with male clients differed from the one they had with female clients. Female actresses, when they did visit Limelight’s offices, usually walked in under cover of sunglasses and proceeded to meet with whomever they had come to meet and then leave. Male actors were much more inclined to relish in the “fun and games” atmosphere that Limelight offered, sometimes visiting Limelight with their 130 friends and hanging out, throwing balls, and joking and chatting with assistants. Thus, although managers had a number of ways in which they could present the emotional setting of their firm, the default emotional tone that Limelight’s partners fostered was the casual one that was geared toward clients (particularly male clients). All of this was done in order to produce a scene that clients and others would recognize as comfortable and legitimate. It was important for talent firms to be seen to clients and potential clients as creatively adept, and it was especially important that they appeared attuned to the popular culture that they were helping to produce. This was reported by several agents at PTA as explaining the ageism seen in the talent industry. The understanding was that one had a window of time in their 20’s to “make it” as an agent or a manager. After that time, if a prospective talent representative did not “have a desk,” then they were considered not in tune with what was current (see Bielby & Bielby 1996 for more on this phenomenon). The imperative to display this sort of casual, hip scene also implies a certain degree of risk. Not maintaining this atmosphere would lead to the perception that the firm was out of touch with Hollywood’s current directions. Ultimately, the fear is that this would mean a loss of clientele. As one PTA agent put it, “you want everything coming from your company to look like it’s ‘fresh.’” On an organizational level, talent firms seek to maintain a hip aura that can radiate through the social networks of the entertainment industry. Generating this aura is one part of the work that talent firms do to produce their reputations. 131 Giftgiving as Reputation Work Sociologists and anthropologists since Mauss, Durkheim, Simmel have noted the importance of gifts in maintaining social relationships (Mauss 1967; Durkheim 1961; Simmel 1950). In the world of Hollywood talent representatives, gifts are given strategically to cement ties with clients and business associates. They are one way in which managers perform “impression management” in front of their clients, and they are an important symbolic means through which a talent representative’s reputation is produced and maintained. In the back stage, however, gifts are sometimes talked about in utilitarian terms (e.g., as “making a good impression”). Managers and agents keep detailed records about their clients’ lives in part so that they can be sure to send birthday, anniversary, and other gifts demanded by a particular occasion. At Limelight, the month of December becomes a frantic time as managers and their assistants prepare an expensive array of gifts throughout the office. Proper giftgiving helps talent representatives put forth the “front” that they are attentive to those with whom they work. There are important institutions of Hollywood giftgiving, however, that talent representatives must enact on the front stage in order to appear reputable. First, gifts must appear to be creative and unique. Limelight gives a large number of gifts throughout the year and maintains a list of stores from whom it regularly buys gifts. Items from large chain stores are generally not given, nor are gift cards. Instead, Limelight’s list consists of a number of “boutique” stores in several cities that were important to their business, including Beverly Hills, New York, Toronto, and Vancouver. These stores specialized in custom baked goods, gift baskets, flower arrangements, or toys. A brochure from one store discussed the sorts of goods that were typical of gifts 132 sold from stores on Limelight’s list of boutiques: “From English Biscuits to Italian Truffle Butter, Japanese Teas to Aztec hot chocolate, and, of course, all the best from here in the U.S., our products are second to none.” Such gifts were sent at various times throughout the year to congratulate clients on some professional success or thank collaborators on particular projects (e.g., agencies, production companies, studio liaisons). Second, talent representatives take pains to make sure that gifts appear “personalized.” For important clients, gifts must convey some knowledge of the client’s interests. For example, one agent at PTA told of how, a few years prior, the company was trying to woo a major director to sign with the agency. Agents at the company had found out that the director had a large art collection, and arranged to buy an original Picasso (reportedly worth tens of thousands of dollars). The gift apparently was successful at opening negotiations, and the client was eventually signed. At Limelight, the partners’ season tickets to the Los Angeles Dodgers were given away on a number of occasions to clients who were also baseball fans. Thus, gifts had to seem appropriately thoughtful. Beyond choosing the right gift, talent representatives can make a favorable impression by the way in which gifts are presented, by giving them the right “touch.” This also can involve considerable back stage reputation work. During Limelight’s Holiday season, managers paid a great deal of attention to the details of particular gifts. There had to be elegant handwritten signatures on the cards accompanying the gifts, labels had to be carefully colored, and presents had to be impeccably wrapped. This was so despite the fact that most gifts were prepared by managers and their assistants dispassionately, with an overall lack of warmth toward the process. Managers overseeing 133 the assembly of gift bags and boxes approached it as something that had to be managed, as something to be done efficiently, against a deadline. Assistants endured it with a sense of exasperation. As Tom’s assistant Mary exclaimed toward the end of one day of Holiday gift preparation, “I fucking hate Christmas!” At Limelight, holiday gifts were given to wide range of people and entities, at a cost that was in the thousands of dollars. There were gift baskets for wealthy celebrities, boxes of gourmet chocolates for agencies, expensive bottles of liquor for writers. It was not a question of whether the recipients needed or even wanted any of these items. Instead, gifts were meant to convey symbolically that the recipients were being attended to, that they were not forgotten during the Holidays nor in Limelight’s business during the rest of the year. More broadly, the gifts conveyed that Limelight was a reputable talent firm, the kind that is successful enough to give expensive gifts. Giving unique, personalized gifts was expected of Limelight, as it was expected of other firms. But meeting this institutionalized expectation simultaneously exposed a variety of risks, felt in the course of the backstage as reputation work was being performed. Did the bags look too “Christmassy” for “our Jewish writers?” Did the signatures have big enough and elegant enough swoops? Were there any price tags left on the gifts? “That would be humiliating,” said Lori, who was helping to oversee the holiday gift assemblage. And, most importantly, did we forget anyone? During the Holiday season of 2005, the risks of improper giftgiving were illustrated by CAA, who had given a custom made digital camera widely to some of its less intimate business associates. The camera was noted in the trade presses as being exceptionally cheap, given that it did not have a zoom feature and had a very small memory. The incident illustrated how giving gifts was something 134 that managers had to do to maintain a positive reputation, but simultaneously opened risks should the gift be poorly received. Reputation Work as Work on the Self The talent representation business is above all about navigating social networks, reading negotiations, and convincing clients and potential collaborators. Given this, it is not surprising that talent representatives perform a great deal of back stage work on their selves in order to appear reputable in their front stage interactions with networks of business contacts. So, what is a “reputable self” in the Hollywood talent industry? Broadly, talent representatives are expected to present themselves as confident, socially adept, and charismatic. When agents and managers spoke at PTA, most had a knack for telling stories and being engaging speakers. Managers at Limelight shared these skills as well. In an immediate sense, these attributes help to make good impressions on clients and other contacts. But they also help more broadly in constructing the perception of a particular talent representative as an “effective person.” As one agent speaking at PTA put it, “to do this, you have to sell your entire person.” Talent representatives often perceived that word about their interpersonal characteristics circulated within their business networks. Furthermore, interpersonal characteristics are frequently cited as key factors by talent representatives recounting how they first moved past being an assistant and “got a desk.” One manager recounted the Catch-22 situation he faced when trying to break into the talent business, and how he ultimately overcame it by presenting himself as a reputable manager should: When I first started, I had no clients and no idea how to get clients. I had no idea how to make deals or anything. And getting clients often depends 135 on having a reputation, which in turn depends on already having clients…. So you more or less have to create the perception that you can do it. You have to have confidence in yourself and show enthusiasm for your projects. Having a confident demeanor was a must for aspiring talent representatives. Along with “networking” and “hard work,” it was a major factor to which talent representatives retrospectively attributed their success. According to several accounts, it helped aspiring talent representatives get noticed by their superiors and, in some instances, directly by clients. In this way, the performance of reputable selfhood is simultaneously the performance of what Goffman (1967) calls the “settled self.” Accordingly, the individual recognizes the rules of particular situations, but is not so aware of them that he or she faces apprehension about performing them (259). To be flustered is to fail to maintain the institutionalized expectation for how to carry oneself as a talent representative. Reputable talent representatives are therefore expected to display a particular set of interpersonal characteristics. However, the precise definition of these reputable interpersonal characteristics varies along gender lines. In a process of sex-typing that is reminiscent of the findings of many sociological studies of gender and work, successful heterosexual male talent representatives are widely perceived as having different characteristics than are successful women or non-heterosexual men in the same industry (see Pierce 1995; Williams 1989). The notion that these differences existed was made explicit during a class held at PTA. A young female agent was a guest speaker, and midway through the class, she made the assertion that women and gay men made better agents than straight men. “We are less constrained. We can see things in a different way. And we can say things that men can’t say.” The course instructor smiled an astonished 136 smile and voiced his disagreement: “I think they make great agents. But I think men make great agents too.” Whatever the disagreement, this perception of gendered differences among talent representatives coincides with a field that reproduces evident gender inequality. Among the firms I visited, assistants appeared roughly split between women and men. However, a sizable majority of full talent representatives are men, as are most of the industry’s most powerful agents. This pattern was true at Limelight, where all of the company’s four partners were men in their forties. In addition, there were two junior male managers and two junior female managers, all also in their forties. Yet, Limelight’s eight assistants (all twentysomethings) consisted of five women and three men. Even if women and gay men did have unique abilities, they still remained largely underrepresented in a stereotypically masculine field. 7 During the course of my fieldwork, I did observe real differences between the ways in which successful heterosexual men and women performed reputable selfhoods. Heterosexual men attempted to give reputable front stage performances of their selves according to several institutionalized, gendered expectations, which included being capable of interpersonal aggressiveness, displaying ownership of physical space, and having a confident sort of swagger (a “gutsiness”) that at times seemed unabashedly cocky. My course instructor at PTA said early on during our first class that “being a successful agent is all about making your presence felt.” And he often admiringly described other male agents who were the subject of class projects as “gutsy” men who just seemed to “make it happen.” One individual whom our instructor spoke of in these 7 The stereotypically masculine nature of the talent industry was often assumed in many explanations of the talent business that were given at PTA. For instance, our course instructor said that young agents were discouraged from attempting to sign “the girl you met last night at the bar” because the agency typically couldn’t “do anything with them.” 137 terms was Tom Young. If being successful was being decisive and “gutsy,” then Tom, who made the move of co-founding his own company, was a good example, according to our instructor. Tom himself described his own outlook during my first meeting with him, saying, “I don’t trust other people to do things better than me. You know, that may sound kind of arrogant, and I say that somewhat tongue-in-cheek. But, basically, it’s how I approach business.” This confident air underlied many of the interactions that male managers at Limelight performed in their everyday work, and these interactions were frequently aggressive in tone. In their worldview, this stance was born of necessity as one had to protect his clients, argue on his clients’ behalf, and ultimately defend his livelihood. Tom recounted an incident to me in which he dealt with one of the more dire risks that can confront talent representatives, which is when another agent or manager tries to poach a major client. I recorded Tom’s description of this in my notes: Tom said that this other guy was befriending his client and telling her worrisome stuff about her future. He said that this guy was there everyday, and that “this was really starting to get to her.” Tom then said that he caught wind of this and called the other manager who was trying to steal his client. He recounted the conversation as follows: Tom: <Flatly> “Hey, how’s it goin’?” Poaching manager: “Good…” Tom: “Here’s the deal: get a new friend.” This aggressive tact repeated itself on a daily basis, as Limelight’s senior managers paced around the office space with golf clubs in hand and talked assertively—sometimes yelling—into their telephone headsets. Accusations were direct and forceful, and the specifics depended on the conversation. Business “was not being done properly,” information was being withheld (“I don’t like being kept in the dark on this”), or other 138 parties were not fulfilling their sides of business arrangements (“This is bad work on the studio’s end”). It was never completely clear to me the extent to which this aggressive front stage presentational style was itself responsible for actually attracting clients. However, some male talent representatives are known to openly advertise their capability for aggressiveness. As one example, George Shapiro (a well known talent manager), used to have his assistant answer his office phone by saying, “George Shapiro's office—kill for the love of killing!” (Friend 2005). At Limelight, not all male managers were quite so aggressive in their interactional styles. Rather, aggressiveness seemed to be correlated with one’s status in the company, with Limelight’s lower level male managers behaving less aggressively. In this way, interactional style became a sort of dividing line between Limelight’s partners and less senior male managers. On one occasion, Tom joked to an assistant and I that Ben, a less senior Limelight manager who had a more sedate demeanor, was “a pussy.” Ability to enact an aggressive interpersonal demeanor thus seemed to establish a hierarchy of masculine performance at Limelight, with some kinds of performances cast as more legitimate than others. At Limelight, reputable masculine performances extended beyond how male managers behaved in the front stage of their interactional worlds (i.e., while negotiating deals for clients) to how they carried themselves in the back stage spaces of their work. Certainly, the playing of sports around the office was part of cementing a common vision of masculinity, as was their frequent attendance together at professional sporting events. At one point, there was talk of holding a corporate retreat at a company that specialized in providing “extreme sports and activities” themed retreats. I was responsible for filing 139 a brochure from this company, which listed some of the possible activities, including bullfighting, sumo wresting, and jousting. Limelight’s co-founder, Dan, had highlighted these activities, and in a notation next to it he had written, “We have to check this out!” Additionally, part of the way in which senior Limelight managers performed their masculinity appeared as though it was meant to convey ownership and control over the physical space in which they worked (possibly as part of broader effort of trying to feel as though they had control over their psychological spaces). Dan illustrated this in one particularly colorful instance when, on a whim, he decided to cut a rectangular hole through his assistant’s drywall cubicle using a small handsaw. His assistant had recently moved cubicles, and Dan explained to laughing onlookers (a screenwriter, other managers, and assistants) that he “wanted to be able to see her” from his office and was thus cutting a window. In contrast, and although I had less of a chance to observe them, the female managers were much more understated in their demeanor. Tina and Lori, the two female managers at Limelight (neither of them partners at the firm) never left their offices while performing negotiations, never paced through the office while talking on telephone headsets, and never played golf on the office putting green. Additionally, I was able to observe both female managers as they talked in nearby open-door offices, and they were far less aggressive while talking on their phones. In my entire time observing at Limelight, I only observed one instance in which a female manager raised her voice over the phone. Institutionalized expectations for female managers were different that they were for men, and female managers at Limelight and elsewhere worked to meet these expectations. Overall, female talent representatives strove to maintain a confident, yet 140 warm interpersonal style. One female agent who spoke at PTA said that her style was characterized by “the BFF approach.” In other words, she said, “make everyone you meet feel like they’re your ‘Best Friend Forever.’” This strategy echoes one that Pierce (1995) found in her study of attorneys, where successful female lawyers often attributed their success to their ability to form deeper relationships with clients. Female talent representatives who did succeed in a stereotypically masculine field were expected to do so by offering more of these “nurturing” skills, and less of the harder interpersonal tactics characteristic of their male counterparts. Therefore, giving a front stage performance as a reputable manager is also a gender performance (see West & Zimmerman 1987). Talent representatives “do” gender, and do it according to institutionalized, gendered expectations for how talent representatives should perform interpersonally. Thus, women and men in the representation business work to establish their reputations, but must do so in fundamentally different ways. Much of presenting this gendered, reputable self to others in the talent industry is an emotional performance. One must be able to produce aggressiveness as a man or nurturing, caring expressions as a woman. In Goffman’s terms, women and men are differentially socialized to figure out “what can be expressively wrung” from given situations, meaning that effective interactional strategies for influencing audiences may vary across sex (1977: 324). In this light, talent representatives strive to adopt the institutionalized gender expectations found in the talent industry. Yet, given an office environment where many social activities seemed to center around sports in which only men participated and where office talk could, without much warning, take on a casual misogynistic tone, the institutionalized ways of performing as a 141 reputable female talent manager at Limelight were disadvantaged when compared with reputable ways of performing as a man. As Schacht (1996) points out, the casual verbal misogyny like that found at Limelight works to not only legitimate more aggressive, confrontational forms of masculinity, but also denigrates attributes commonly associated with femininity. Female and male talent representatives might have been able to “expressively wring” in different ways from the situations they encountered as managers, but this still translated into evident gender inequality when it came status within the company. For both female and male talent representatives, however, perhaps the most important performance is being able to project a confident front throughout a stressful job in which one’s standing depends on the perceptions of clients and other contacts. Yet, maintaining this reputable confidence and poise requires talent representatives to emotionally prepare themselves in the back stage—what Arlie Russell Hochschild calls “emotion work” (1983, 1979). In the Hollywood talent industry, this emotion work is an attempt to impose a sense of order upon a self that can veer toward instability (see Giddens 1991; Gergen 1991). David Wirtschafter, who heads William Morris Agency, put it as follows: “Doing this job is about removing emotion, about creating calm. Calm creates order, and order allows you to solve problems” (quoted in Friend 2005:53). This search for “order” explained much of how managers at Limelight prepared for their work. Limelight’s sleek, carefully planned physical aesthetic was part of this. In one sense, it was an attempt to meet an industry-wide institution for how talent firms should physically look; but it also helped create a suitable physical environment for the back stage emotional labor that had to be done by managers. The “building consultant” hired 142 by Limelight’s partners was part of this quest for emotional order. She came through the office on a monthly basis to work on the office’s “energy.” On the day I observed her visit, she carried a metal pot containing burning incense and alcohol that produced a large flame (approximately two feet tall). She moved from office to office and cubicle to cubicle, holding the pot over every chair, corner, and desk. When I asked her what was in the pot, she told me the ingredients and said that it was “good for getting out old, crusty, sitting-there-all-day-long energy.” Everyone in the office accepted the consultant’s presence as an unremarkable occurrence, and one of the partners happily moved to allow her to work in his office, saying, “Bring it on.” I observed other attempts by managers to impose order on their chaotic emotional worlds. For example, Limelight’s partners hired a full-time personal assistant to help them manage the demands of their home lives. One partner had regularly scheduled “stress tests.” Another had a fairly unconventional concern with his health, and regularly had special blood tests done to measure his metabolism as well as quantities of various vitamins in his body. Despite their attempts at maintaining emotional order and calm, such a state did not always hold for Limelight’s managers. Breaks in this order could occur at almost any time, and surprises were just a phone call away. Relationships could suddenly become strained with clients and clients could threaten to leave. One such incident came when Tom Young was involving a major client in negotiations and had briefly put her on hold to talk with someone on another line. When he switched back, he soon realized that his client had hung up on him. Tom suddenly grew very frustrated, and angrily shouted to his assistant, “Where the fuck is she?!” He seemed to think for a moment over his 143 assistant’s quiet explanations, and then continued: “She’s a bitch! All we’re trying to do is get her a job. She’s a bitch…” Frustrating moments like this were fairly common at Limelight, and were significant because, although a manager might vent to an assistant in the back stage, the deal making had to continue on the front stage. I observed on a number of occasions Tom and other Limelight managers attempting to maintain the flow of business after these “blowups,” providing a strained smile while greeting a female client, for example, or uncomfortably taking another phone call that required a calmer tone. One agent speaking at PTA explained the emotional strains of being a talent representative this way: “Hollywood’s a funny place in that the work is never done. And you can let that crush you. There is no ‘agent season.’” Several talent representatives I observed, including Tom Young, said that there was a part of being a talent representative that was captured by a famous “dark metaphor” that circulated in the business. The metaphor held that being a talent representative was like floating in the ocean and facing a constant succession of waves. One can ride one wave and successfully stay afloat. But if one lets his or her guard down, he or she risks being crushed by the next wave. The above metaphor is apt in describing the back stage emotional lives of Hollywood talent representatives. “Floating in the ocean” captures a sense of isolation that stresses how the individual must fend for him or herself. The looming waves allude to the dread and perception of risk one faces if he or she fails to focus and maintain a “settled,” reputable selfhood. The work that it takes to display this reputable self highlight the risks that talent representatives perceive should they fail to maintain their expected confidence. 144 Overall, these various forms of “reputation work” highlight how Limelight’s managers swam in an interpersonal culture that was dominated by informal conventions that governed how one achieved legitimacy within the business networks of Hollywood. It was in this context that intuitive discourses about audiences were used in judging movie ideas. How Movie Ideas Were Evaluated at Limelight Within this world of reputation building, talent representatives make decisions about movies. Much of the work of talent agents and managers involves activities that have little directly to do with the evaluation of movie ideas, but these activities nevertheless form the context within which movies are evaluated. Given their current position in Hollywood’s organizational ecology, talent firms occupy an important place in which decisive evaluations of movie ideas occur—where intuitive discourses about movies are spoken and intuitive knowledge about audiences is recreated. Limelight Entertainment provided a good site from which to see these intuitive discourses in action. In addition to being a talent management company, Limelight also had a small production wing that was attempting to put together full motion picture projects for potential studio buyers. Thus, at Limelight I was able to see how movie ideas were evaluated as prospective projects to be produced as well as how movie ideas were presented to other entities outside the company. Overall, then, I was able to observe how managers at Limelight talked about and presented movie ideas using their intuitive senses of what would appeal to audiences. 145 One way in which I was able to see larger patterns in how movies were talked about at Limelight was through access to various company documents. Among the documents that provided clues to how managers talked about movie ideas was a “coverage summary” that contained managers’ brief evaluations of 297 screenplays. These scripts, I learned later, had been submitted to Limelight’s production wing over a period of several months when they were looking toward putting together a new motion picture project. The scripts had been covered by various managers (or sometimes their assistants) at the company and summarized in the document. Each summary in the document contained the title of each script, the names of its writer(s), the name of the writers’ agent, a very short (2 sentence or less) plot synopsis and evaluation. In this sense, this coverage was different than the more extensive coverage done for studio executives by paid analysts (such as those I interviewed). At Limelight, managers/ producers did coverage themselves for their own internal use. Of the 297 screenplays reviewed by Limelight’s managers and their assistants for its production wing, only 27 (approximately 9%) received a final evaluation that was something other than a “pass.” In these few cases, notations on the report summary indicated that Limelight’s managers would arrange an in-person meeting with writers and advance the movie idea to the next level in the process. Although most of the managers’ comments were very brief, many did provide clear evaluative statements, the vast majority of which were negative in some way and ultimately indicated that a movie idea had been removed from consideration. A majority of evaluations, however, provided no clear reason why a movie idea was rejected or advanced. In fact, most made a simple, non-specific statement such as “not good” or “seemed okay.” However, a sizable number 146 of scripts—114 of them—received specific evaluations that gave clear reasons why they were “passed” on (these reasons are summarized in Table 4). ________________________________________________________________________ Table 6: Reasons given in Limelight’s script coverage summary for “passing” on screenplays ________________________________________________________________________ Evaluations of writing mechanics, quality 9 “Dialogue felt forced,” “dialogue off,” “dialogue weak” 3 “choppy” 1 “lame dialogue” 1 “dialogue cheesy” 1 “not well executed” 1 Unrealistic characters 1 “not much characterization” 1 Evaluations of writers themselves 26 “not enough/very little experience” 25 or “credits iffy” writer has a “bad reputation” 1 Evaluations of stories 79 “boring,” “not interesting,” “dull” 22 “not funny” 8 “dark” 7 “Cheesy” 4 “no story”, “story seemed off” 4 “uninspired,” “unoriginal” 4 “Yuck” 4 “juvenile,” “silly” 4 147 Table 6, Continued “odd” 3 “weird” 3 “too sad” 3 “Over the top” 3 “slow” 2 “predictable” 2 “lame” 2 “complex” 1 “dated” 1 “soapy” 1 “Not plausible” 1 Table 6 shows reasons why movies were evaluated negatively. Consistent with the findings from my interviews, both perceptions of writing mechanics and writers themselves did serve as stated reasons why scripts were rejected. Accordingly, writing— especially dialogue—could feel “forced,” “choppy,” or not well-executed in general. In addition, the experience of prospective writers was spoken about as an important factor in a sizable number of evaluations. In some of these cases, Limelight’s managers gave no indication that they disliked the story idea itself and even spoke positively of it. However, a writer’s lack of previous credits was sometimes used as a quick justification for disqualifying a script. The most frequent kind of justification for disqualifying scripts had to do with managers’ perceptions of the movie ideas themselves. Limelight’s production wing was interested in making some sort of more “commercial” comedy, “dramedy,” or action 148 feature. Given this, movie ideas were sometimes negatively evaluated using a language of how they were not “commercial” (in various ways). The scripts that most obviously seemed outside the realm of being “commercial” included those that were “too sad,” “dark,” “weird,” “odd,” or “complex.” Some movies were simply described with a “yuck,” which seemed to indicate their exceptional strangeness, given the conventions of “commercial” Hollywood movies. For example, one such movie idea in this vein involved a story that would have been told to a great degree by the main characters speaking through “thought bubbles” (like the bubbles that appear in comics). These various descriptors hinted at how movie ideas such as this lay somewhere outside the bounds of “entertainment”—at least the kind that would appeal to broad audiences. Other movie ideas were evaluated in ways that suggested that they were not “edgy” enough to be movies that were sufficiently “commercial” for Limelight’s production purposes. These movies were “cheesy,” “juvenile,” and “silly.” Or, they were spoken of as too inauthentic, as “not plausible” or as “over the top.” Various other evaluations may have hinted at a lack of edginess, including the many scripts that were labeled as “boring” or “unoriginal” (even though these were somewhat different ideas). These were marked as lacking any sort of fresh idea, having stories that were too similar to movies that had already been done. The script that one of Limelight’s managers called “dated” also lost edge by seeming locked in another era—the 1980’s, the era in which the writer who submitted his screenplay had his last credit. The predominant way in which managers at Limelight evaluated screenplays, then, drew on familiar intuitive discourses that, given what I found during the interviews I conducted, circulated elsewhere in Hollywood. The intuitive categories of 149 “commercial,” “artistic,” and “edgy” fundamentally structured how managers at Limelight understood movies. They provided larger frames of reference for understanding types of movie ideas, and served as background for the dealmaking that had to be conducted on a daily basis. During an interview, one of Limelight’s partners (Dan) told me how his understanding of the difference between “commerce” and “art” was a core one in how he had eventually come to understand the business. Dan was an avid poker player, and described this disctinction in those terms: I seemed to have a way with the actors. At least I understood them, I appreciated them. And I understood pretty quickly the difficulties of being an artist in the business. And the difference between art and commerce, and how difficult it is for people to sort of rationalize what they’re doing and still call it art and still sort of want something to be commercially successful. It’s a struggle that I think goes on with all artists in this business. Because, ultimately, you know, it’s like saying you love to play poker. I just love to play poker, all I want to do is play poker, that’s all I care about is playing the game, I love the game. But, being in Hollywood is like being in Vegas playing poker at the Bellagio. Wait, so if you just loved playing the game, why didn’t you stay in New Jersey and play your Friday night game with your buddies? Why did you come to Vegas and play at the Bellagio? Because you want something else, you want something a little more commercial. You want a bigger success or a bigger win or whatever it is. For Dan, this “commercial” versus “artistic” dichotomy was a taken-for-granted way of understanding how Hollywood movies were to be understood. These categories shaped how Dan and his clients pursued particular movie projects. In my interview with him, when I asked him about how he selected scripts for his biggest A-list client (a well known female action star), he offered the following response: Um, well, the last thing that she wanted to do was, meaning the thing she wanted to do the least, was a tiny little independent movie for no money. Several months ago. She was in a state where she had just done a couple of them and she was like, you know what, she just felt like—we all felt like—it was time to take advantage of a commercial movie. And we had 150 the benefit of knowing that she was going to go do the sequel to [a major action movie] because she had already had a deal in place, and all that. So, when she finishes that big commercial movie of just doing silly dialogue and, you know, stuff that makes that movie work, a lot of action, a lot of sort of fun but goofy dialogue, and all that, and then goes all over the world selling it, talking about it, and really realizing it’s sort of global impact and it’s sort of, you know, enormous reach, she comes out of it thinking, “God, all I wanna do is sort of escape into a little movie.” One that’s off the radar, there’s a great acting experience, where I’m not hoisted up on cables every day, or I’m not standing in front of a green screen, or whatever. And, there’s a movie that we’re talking about right now that’s, several months ago had it been submitted, I would have been like, “This isn’t one she’s looking to do. She’s not going to do this one. It’s a ‘pass.’” But now it’s exactly what she wants to do. “A lot of action,” “fun, silly dialogue”—these are things that make movies “commercial.” And, although Dan saw this acting client as able to decide between pursuing either a “tiny little independent movie” or a bigger “commercial” movie, this line between “commercial” and “artistic” was a meaningful one that described the characteristics of different kinds of movies. From the perspective of Dan’s acting client, “commercial” movies are ones where one cannot truly create, or look to satisfy some larger existential desire. Commercial movies were about making commodities, pure and simple, and not vehicles for “a great acting experience.” Conclusion: Intuitive Knowledge Reproduced in the Context of Reputational Games Hollywood, as a site of decision-making and a place in which knowledge about motion picture audiences is produced, is not a uniform place. Multiple forms of knowledge about audiences coexist and are reproduced within different organizational conditions. In this chapter, I have attempted to describe how one form of knowledge about movie audiences—in the form of intuitive discourse—is produced within the 151 milieu of the talent industry. Here, I have shown how “intuitive discourse” works as a set of institutionalized ways of speaking about audiences. Notions of “commercial,” “artistic,” and “edgy” become meaningful conventions that moviemakers use in describing movies. In addition, such discussions take place within an interpersonal world of dealmaking that is dominated by important informal conventions for how to relate to business associates. What I argue here (and differently in the next chapter) is that these two realities are intimately related. Some sociological studies on the “culture industries” have shown how cultural producers use institutionalized ways of speaking about cultural products as a way of stabilizing opinions about those products where all opinions would otherwise be held as entirely subjective (see Lampel et al 2000; Ahlkvist & Faulkner 2002; Bielby & Bielby 1994). In other words, rhetorical conventions for talking about TV shows, potential popular music hits, or movie ideas proliferate in places where, as William Goldman put, “Nobody knows anything.” What I argue here is something slightly different. The decisive feature of the Hollywood talent industry, as I see it, is not so much that nobody knows anything about the product, but more that nobody knows anything about relationships. Talent representatives I observed and spoke with were most concerned with issues of maintaining their positions of legitimacy among their peers. Relationships with clients, studio executives, and other business associates were formed within a hierarchical organizational field that legitimated based on the collective perceptions of companies and individuals. This fact helped define the lived texture of being a talent representative, one where “reputation work” was performed to help meet the challenges of flux within the industry (where companies and individuals can rise and fall very fast) 152 and of trusting one’s associates when information about deals was of paramount importance. Within this context, intuitive ways of talking about audiences became the taken- for-granted language of choice for those working in this segment of the Hollywood film industry. This language did provide stable frames of reference for talking about cultural products (in this case, potential motion picture ideas). But, from the standpoint of providing a stable form of evaluating products, there did exist better possible languages. Specifically, a readily available repertoire for evaluating movie ideas in Hollywood involves a technical language of demographics, audience testing, and predictive “scientific” measures (discussed in the next chapter). As presented in Chapter 4, this technical knowledge offers a less fragmented, less malleable way of talking about audiences. Although the measures and fundamental categories that form the basis for this technical language are socially constructed, they are ultimately accepted as firm and definite by industry participants. Given this, why don’t talent representatives (and many studio executives) emphasize technical forms of evaluation over intuitive ones? In the few cases I observed “technical” information being discussed at Limelight, it was treated not as managers’ evaluative frame of choice but as a barrier that had to be dealt with. One instance of this involved an actress for a prospective television show that Limelight’s production wing was putting together. One of the show’s writers (a Limelight client) and two of Limelight’s managers were discussing a prospective actress who was to potentially work on the show. The writer was remarking how the actress “looked really good today” and how she was “smart and funny.” One of the managers/ producers agreed, saying, “Yeah we have to get her.” The “hitch,” as one of the 153 managers then reminded the other two, was that “she didn’t test well… She tested too old.” Confronted with this fact, the three of them began discussing how they might go about having her “retested,” given how much they all liked her personally and liked her as an actress for the part they were casting. In this instance, then, Limelight’s managers did not appear to put a great deal of stock in “objective” audience data as a tool in making decisions about the product they were seeking to produce. Dan, a partner at Limelight, spoke to this issue when I interviewed him. As he put it, It’s like, in our line of work, we don’t do market research to see what new fads are or what kids want to see next, or any of that kind of thing. We hear about stuff like that, you know. But it doesn’t really effect the management business as much. Thus, technical knowledge about movie audiences did not much help Limelight’s managers in making decisions about movies, but depending on its conclusions, it could either hinder or help advance potential movie projects. 8 When we consider how technical audience data is not emphasized in the talent industry, it becomes easier to see why intuitive knowledge is drawn upon more heavily. Although intuitive categories such as “commercial,” “artistic,” and “edgy” do appear to have shared meaning for moviemakers, these categories are ultimately more malleable than technical evidence about motion picture audiences. As Chapter 4 discusses, technical measures about audiences, once they are in place and taken as “real” by industry participants, are simply held as objective facts that have to be confronted. When some product or some element of some product appeals to a certain percentage of 12-17 8 In terms of how technical data could potentially help advance a manager’s agenda for a product they were trying to package and sell, I did see correspondence between Limelight’s managers and those outside the company regarding the same television show discussed above (in which the desired actress had received bad test results). But, here, the managers mentioned how the show had “tested very well,” suggesting how technical audience knowledge became either convenient or inconvenient for advancing a movie, but not a criteria very much considered when making decisions about content. 154 year olds (according to some measure), this can either work to favor or hinder a movie idea’s advancement in the Hollywood system. But this “fact,” such as it is, cannot be fundamentally changed—regardless of one’s position in Hollywood’s business hierarchy. This is precisely part of why scientific audience research was greeted with such skepticism by many top studio executives when it first appeared in Hollywood in the 1930’s (see Ohmer 2006). On the other hand, when movie ideas are discussed in terms of them being “commercial,” “artistic,” or “edgy,” they are spoken of in evaluative terms that carry weight, but that are ultimately more contestable. Intuitive evaluations are much more intimately tied to the speaker’s own interpersonal legitimacy, and reputations can be staked on these judgments. As Dan put it when I interviewed him, “There are no rules. If you’re good, you’re good, and if you’re picking good material, you’re better.” In this way, intuitive languages for evaluating motion picture ideas thrive in exactly the sort of organizational context described above—one rife with reputational games. Technical evidence can either support or embarrass, but intuitive judgments are matters of perception constructed amid variously legitimate players within Hollywood’s dense business networks. Speaking about movies in this way allows individuals and firms to more readily hitch their reputational clout to potential movie ideas. In this way, intuitive evaluations are made possible within organizational conditions in which individuals and firms jockey for perceived legitimacy. 155 Chapter 4: Technical Knowledge in Reproduction “The stuff works.” –Eric, CEO of a major entertainment research firm, referring to the methods used in movie research In 2006, the Motion Picture Association of America (MPAA)—the trade organization representing Hollywood’s major motion picture studios—made public the results of its annual “U.S. Movie Attendance Study,” which provides a statistical snapshot of American moviegoers (as measured by the research firm commissioned for the study). The short, eleven page report displayed several facts and trends regarding the moviegoing public in 2005 that were thought to be important background for those involved in the American entertainment industries. Among the report’s key findings were that young moviegoers (those ages 12-29) continue to represent the bulk of moviegoers; that there was a relative shift toward younger moviegoers in 2005, as evidenced by increasing attendance from the 12-24 year old demographic and declining attendance by those over 50; and that a disproportionate number of “frequent” moviegoers are also young (MPAA 2006). In 2007, the MPAA released its next annual “U.S. Movie Attendance Study” using data on 2006 attendance. Of the 175 million U.S. moviegoers who bought 1.3 billion tickets, the vast majority continued to be “young”— even though there was “moderate” growth (measured in ticks of percentage points) in attendance among older groups (see MPAA 2007). Overall, the reports in both years seemed to present a resounding emphasis for those connected or interested in the movie business: “think young.” 156 The MPAA reports are examples of how Hollywood has come to see its target audiences in terms of differentiated age groups. As Chapter 2 discussed, this overpowering awareness of “age” as a differentiator of consumers is a relatively recent one in the film industry, and one that emerged out of mid to late twentieth century debates within the industry over its audiences. If particular demographic categories were regarded with skepticism in the past, they now form a basis for how many in the business think about who will watch what movies. And the idea that the American movie audience is overwhelmingly “young” is but one “fact” in a constellation of knowledge about who movie audiences are and what sorts of films they prefer. “The numbers” allow for a variety of other generalizations about demographic groups, ideas which I heard echoed across a number of sites during the course of my research. They include the ideas that “independent” films resonate with relatively affluent, well-educated white urbanites from “blue states;” that teenage boys “from the Midwest” constituted the most typical piece of the special-effects hungry blockbuster audience; or that people over 50 do not go to movies with nearly the intensity as their younger counterparts. Together, such statements form a map covering much of the active and potential moviegoing populations. I call these audience images demographic types because they paint a normative picture for what kinds of films particular audience segments prefer and how they prefer to watch them. On the face of it, and going by available quantitative data, all of these demographic types would appear be somewhat accurate. Yet, as I show below, there are reasons to doubt the validity of many proclamations made by commercial audience research. Audience research in the movie industry is an imperfect process, and is 157 sometimes acknowledged as such by those familiar with the industry (see Marich 2005). It shares commonality with academic sciences that lay claim to the notion of “objectivity,” but whose circumstances of production are often at odds with the possibility of pure objectivity. Sometimes, as some social studies of science have pointed out, the ways in which scientists interpret information to make knowledge becomes a strategic game—one that is intentionally spun to produce particular kinds of stories (Hilgartner 2000). At other times, the plot of knowledge production appears less insidious, though scientists nevertheless routinize their work in such a way that the research process fails to question key assumptions or even favors particular kinds of conclusions (Latour & Woolgar 1979). Overall, then, science is produced in local contexts that are often at odds with pure objectivity (Knorr-Cetina 1999). 9 And, although commercial audience researchers do not purport to be “scientists” and are not regarded within the film industry as such, they are no different on this point. Audience research appeals to a core principle that data being gathered and interpreted is “objective,” and audience research firms trade on this principle as well. Similarly, Hollywood decisionmakers are able to legitimately cite audience data precisely because of this presumption of “objectivity.” However, the production of technical audience knowledge is situated in a process that is uneven and questionable from the standpoint of “objectivity.” The industry’s final numbers on audiences, presented in flashy Powerpoint reports, obscure that process. 9 This also bears some resemblance to several classic studies of journalism, which reveal how journalists—while claiming “objectivity”—nevertheless work within an epistemological framework that limits the types of news narratives that can be written (see Tuchman 1978). 158 Second, the truth value of demographic types, now operating as kinds of folk wisdom within the movie industry, is further complicated by the fact that these demographic types may help bring about the very truths that they proclaim. This is because these broad audience “truths” constitute shared explanations that Hollywood decision-makers use in their work, guiding them to conclude that certain kinds of films are feasible while others are not. Take, for instance, the major film studios’ approach to making movies for the “youth” market. If last year’s data shows that teenagers constituted the largest relative group of moviegoers and the most frequent group of moviegoers and spent the most on movie tickets, then it would seem to make a certain kind of sense to continue to produce the kinds of movies that young people went to last year, devote a vast majority of marketing resources to those kinds of films, and release those types of movies on as many screens as possible. Young audiences, for their part, may come to pick up on cues (provided by intense marketing that particular kinds of films are made for them), thereby embracing a self-identity that is similar to the industry’s understanding of them—a process some researchers have called interpellation (see Banet-Weiser 2007; Williams 2006). However, this situation introduces a great “chicken and egg” question that lurks beneath the surface of Hollywood’s audience epistemologies. What if movies were made and marketed differently? What if the kinds of small movies that seem to appeal to older moviegoers were widely released and marketed as major pop culture events, in the same way that major summer blockbusters are? If this were to happen, might it actually change what movies are watched by what demographic groups? As it stands, the existence of audience research and the conventional conclusions that those in the industry draw from it may participate in a kind 159 of self-fulfilling prophecy whereby technical ideas about movie audiences actually help make those audiences in the first place. In this chapter, I examine how technical knowledge about movie audiences is made, interpreted, and eventually reproduced. In doing examining the conditions under which this kind of audience knowledge is made, I hope to illuminate how movie audiences are not objective entities whose realities are simply measured by commercial audience researchers. Rather, I suggest that their existence as forms of knowledge is something that commercial audience researchers must continually accomplish (Ang 1991). The work of audience research is done against the backdrop of epistemological assumptions that are not inherently natural, against a research process that researchers themselves see as suspect, and against a research process that does not lend itself to self- reflexive criticism. How this “accomplishment” happens across the various sites of commercial audience research is traced below, from how data is gathered, to how it is interpreted, to how it is presented as evidence for those who use it to make decisions about movies. How Audience Data is Made: Test Screenings and the Movies On a balmy August evening in Southern California, as the sun starts hanging low in the sky and colors shine brilliantly through the hazy air, a line begins forming outside a movie megaplex in a suburb of Los Angeles. The megaplex is near a large mall and outdoor shopping center, and shoppers begin taking notice of the small mass of people that has begun gathering in front of the theater. Several of them, out of human curiosity, walk past the restaurants and street musicians to look at the theater and the amassing 160 crowd to see what is happening at the megaplex and why the crowd has gathered. Over the next two hours, people will continue lining up. Groups of teenagers sprint down the line to claim their spots and try to ensure that they will get in. A man in his 30’s, who is holding a place in the line, says loudly into his cell phone that, “I think they’re going to start letting us in in about 15 minutes… You better get over here!” Other people in line wait for a few minutes and, deciding that it is not worth the wait, walk off. On this evening, the theater is hosting a test screening for a comedy starring several well-known actors. A first cut of the movie has already been made, and the studio releasing the film has hired an audience “research vendor” to conduct the test screening and gauge a live audience’s reaction to the film prior to its release. As people continue to arrive, young men in their early twenties dressed in suits greet the rapidly multiplying crowd with sheets of paper and pencils. Prospective audience members fill out these cards that ask about demographic information, movie consumption habits, as well as contact information to be used in recruitment for future screenings. I go dutifully through the questionnaire, thinking carefully about each response and its accuracy. How many movies have I seen in theaters during the last year? Is it really over 12? I eventually make my best estimate and fill in the proper bubble. Meanwhile, many fellow audience members around me have already finished their questionnaires, some easily within a minute of when they were handed out. I have come to this screening alone, but behind me there is a group of teenagers who are chatting and goofing around, giving only partial attention to their form and intermittently consulting with each other about how to fill it out (“Hey, am I Hispanic or Latino?”). The young men in suits collect the questionnaires, exchange them for small tickets, and, glancing at the questionnaires, 161 make tick marks on their clipboards to get a rough count of how many audience members are male or female and how many are 25 and older versus younger than 25. Soon, the young men in suits start letting groups of twenty or thirty people into the theater and direct them past the snack bar to a large theater to the right. At this juncture, the test audience members will go through security. Three larger men (also in suits) stand behind a folding table, holding hand-held metal detectors and running them quickly over the legs and extended arms of each audience member before they go into the theater. There is a sign posted at the security table reading “NO CELL PHONES WITH CAMERAS.” Fortunately, my cell phone does not have a camera, but I see one woman being told that she cannot enter the theater with her camera-ready cell phone. I know from talking with people working in the industry that fears of movie piracy and the potential for leaks about the movie are ever-present worries surrounding the test screening process. If a movie is recorded beforehand, it could find its way onto the Internet where it could be downloaded for free or even put onto a bootleg DVD sold throughout the world in advance of the official movie’s official release. 10 In the end, however, test screenings are usually considered to be a risks worth taking. Once the security wand passes over me and my hand is stamped, I walk in. The theater itself is a huge, stadium-style theater seating approximately 450 people. I am among the first people to get in, and I sit up near the top, in the middle. Several seats away from me is a large taped off area marked “RESERVED.” This is where “the suits”—the industry professionals who have worked on the movie being tested—will sit. 10 Bootleg copies of movies are significant not only in the United States and Europe, but especially in many Asian countries. In the cases of some countries, it is believed that a sizable majority of Hollywood movies viewed are watched using bootleg copies. 162 People continue filing in, increasingly finding it difficult to sit near those they came in with. Employees of the research company continue to teem about, talking on walkie- talkies and directing people to their seats. A man at the front of the theater works on an infrared camera affixed atop a tripod, which will eventually watch the test audience watching the movie. Some studio executives and directors are said to find such footage valuable. The audience begins to completely fill the theater, and a feeling of expectancy starts to fill the air. After approximately 45 minutes of sitting in our seats, a man in a suit comes to the front and announces that we will get started in 15 minutes. What will the movie be like? Some people continue to get up to use the restroom or get snacks. Some of the “suits” have taken their reserved seats after chatting in the lobby or side aisle. As they do so, they draw curious gazes from a few audience members (What makes them special?). Some of the “suits” are actually wearing suit coats, but some are dressed more casually. Many of them seem indifferent to the people gazing at them, and sit chatting with each other, checking their Blackberries, or eating candy. As promised, the suited master of ceremonies soon returns to the front and draws the crowd to attention. He announces in a showman-like fashion, “You are in for a very special treat tonight. You are one of the first audiences in the country to see this film. Give yourselves a round of applause!” In response to this enthusiastic introduction, most people in the audience applaud and some shout. This is perhaps because they are happy about seeing the movie, perhaps because the movie is finally beginning, perhaps because other people in the audience are applauding, or even perhaps because they want to show excitement for those who are showing the movie. After the applause, the announcer tells 163 us that although most of the film is finished being edited, parts of it may still be a little “rough” and that we should disregard any parts that are discolored. Very importantly, he says, we should remain in our seats after the movie is finished playing, because there will be a short questionnaire to fill out. This is my first time observing a test screening, and, no doubt, this experience has so far proven different than other movies I have seen in theaters. Instead of trading money for a movie ticket, we will all have traded approximately an hour and a half of extra time (counting the time we waited in line, waited in the theater, and the time we will spend filling out the survey after the movie) to see a movie we have only heard about in the vaguest of terms. We are also watching the movie next to people who are connected with the movie, creatively and financially. I feel aware of this, and have the brief impulse of wanting to be complimentary to the people putting on the show. Successive visits as a participant observer at audience test screenings revealed to me that test screenings are their own unique scenes. Observing these scenes repeatedly allowed me to assess how the stated aim of such test screenings matched what actually happened at these settings. How did the test screening process square with the presumption that this was a scientific research method? Audience research, like other “scientific” research, claims legitimacy based on its appeal to a principle of “objectivity.” Fundamental to this claim is the assumption that data collection proceeds in a pure, untainted fashion. This is the basic operating logic of positivistic science (the form of knowledge acquisition that audience research is modeled upon). If we were moviemakers and we were applying this logic to the audience test screening as a method of research, we would have several specific expectations. First, we would assume that 164 data would be gathered in a controlled setting, such that the only factor influencing how an audience member evaluated the movie being tested was the movie itself. Thus, we would assume that an audience member watched and evaluated a test movie in a setting that did not vary substantially from one test screening to the next. That is, because the results of one test screening are assessed based upon how they stack up against the industry’s collected knowledge about other test screenings, there should be nothing different about the setting of any one test screening in particular. Furthermore, because each individual audience member’s opinion counts equally in calculating the audience’s average opinion, we might also have to assume that each audience member’s opinion is equally important in actually forming the opinion of the whole. Ideally, then, no audience member should be able to influence the ultimate evaluation of another audience member. Second, from this vantage point, we would also assume that test audiences are truthful and accurate in the evaluations they have provided us (i.e., our data). If an audience member does not care enough to think about each item and check an accurate answer, then he or she is not providing us with the kind of information we are assuming we are getting. For example, if a test audience member does not read survey items and checks them as fast as she can so that she can leave the theater as quickly as possible, the data she is providing us is completely random and introduces error into the calculations we think we are doing. Third, since a “test audience” is supposed to do that very thing—test, there is an even more problematic assumption that has to be made: that test audiences in general mirror actual movie audiences. Movie theaters represent a context of reception (Srinivas 165 2003), and the context of the test screening ideally should not be any different than the context of a regular trip to the theater. And if it is in fact different, the test screening process should not systematically bias audience members one way or another. That is, in a basic sense, audience members should not feel any differently about a movie they see at a test screening than they would have if they had seen the movie during a conventional trip to the movies. Additionally, test audiences should be representative of movie audiences in general. We might assume that tastes will vary based upon demographic categories, in which case we would want our test audience to have the same demographic makeup as the moviegoing audience at large. However, it might not be practically possible to gain a demographically representative sample, and we might have to instead build demographic quotas into our design (an imperfect solution that audience researchers adopt). Even still, if audience members (regardless of their demographic characteristics) come to test screenings for reasons that are substantially different than the reasons why general audiences pay to go to movies, there is further potential for error. Or, at the worst possible extreme, if test audience members never pay to go to the movies but go to lots of test screenings instead, there may be substantial inaccuracies. Afterall, we are trying to assess how actual, paying consumers will respond to our movie. If we then had to make decisions about the movies we tested (how to market them, how extensively to release them, etc.), we could disconcern ourselves with whether or not test audiences reflected actual audiences and simply view test audiences as abstract indicators that we have found to be correlated with observed patterns in the “real world.” In other words, test audiences wouldn’t really reflect how actual audiences behaved, but they might approximate certain patterns in a useful way, given the types of 166 decisions we concerned ourselves with. For example, we might be able to observe that when a test audience’s average answer to the question “How likely would you be to recommend this movie to a friend?” exceeds a certain threshold, the movie being tested is more likely to play well at a large number of theaters. However, if test audience data becomes too far abstracted from actual audiences as real phenomena, the correlation between survey item and audience behavior might be extremely rough and error-prone. And, perhaps, our surveys might cease to really measure elements of the moviegoing experience that are most central to why people go to movies. All this is to suggest the real world difficulties that are present in conducting audience test screenings. Despite a core claim of “objectivity,” it is extremely difficult for test screenings to satisfy this basic assumption in practice. This is not to necessarily single out the test screening process as a particularly deficient form of data collection. Indeed, other scientific enterprises, to varying degrees, deal with a gap between assumed objectivity and the often less-than-objective process of gathering data—an issue that has been a recent focal point in the philosophy and sociology of science (see Gieryn 1983). Data collection is a process that occurs amid real-world chaos, against routines that are often at odds with the possibility of “objectivity” (in its purest sense) (see Galison 1997; Latour & Woolgar 1979). In the world of movie audience test screenings, knowledge about audiences is made according to a priori biases that are intentionally built into the design of data collection as well as unintended threats to objectivity that emerge during the collection of data. What is especially interesting is that those conducting and analyzing test screenings often seem aware of problematic elements in the data collection that occurs at test screenings. At times, they even appear fully complicit in building 167 “objectivity problems” into the process. The biases of the test screening process are described in more detail below, for they are a first stage in formulating an overall scientific picture of movie audiences. As such, they are also the bases for the Hollywood film industry’s technical language for describing its audiences. Audience Test Screenings in Industry Context Audience test screenings can be carried out either directly by motion picture studios or one of a handful of independent “research vendors” that service various companies in Hollywood. For an approximate cost of about $20,000, a studio can conduct its own test screening and get the kind of information it provides. In a class I attended on movie marketing (taught by Mark, a former chief marketing executive at a major studio and a key informant on the research process as a whole), one business student expressed disbelief at the cost of doing a test screening. This seemed to be an exceptional amount of money for one piece of research. However, as the marketing executive pointed out, “You spend millions of dollars on a movie. This is really peanuts.” Test screenings have become a regular part of the movie making process. Although their nature and use has changed over their history in the film industry (Ohmer 2006; Handel 1950), they are now most regularly done after a film is mostly complete. Their main purpose, according to those working in this area of the business, is to assess both a film’s “playability” and “marketability.” The notion of playability refers to how well an audience “likes” a film, while marketability refers to elements of the film that “create value.” In other words, as key concepts in the field of movie research, 168 “playability” covers audience preferences while “marketability” covers what is sellable to those audiences. Accordingly, measurements of a film’s playability are thought to be useful in determining how popular a film will be. These measurements may aid decisions about how widely to release a movie (if the movie is to be released in theaters at all) and help make predictions about what a film will make. On the other hand, measurements of a film’s marketability are seen as useful in deciding how to frame a film concept for potential consumers. Elements of a film seen as marketable become important as the studio develops and targets its advertising for the film, which will come in the form of television spots, radio ads, Internet promotions, and theatrical trailers. Beyond playability and marketability, a related but equally important concept is that of “the target audience.” This refers to the category or categories of movie consumers the film will be aimed at. In audience testing, these categories are most often marked by the demographic identities of consumers. Categories such as age and gender are considered to be particularly important, followed by education (as a kind of measure of social class), then region, urbanity/rurality, and race (all of which sometimes stand in for social class as well). Ideally for the studio, a film should be especially playable among its target audience. And, its marketable elements should be those that are thought to appeal to a particular demographic target, primarily because it is these marketable elements that will be used in advertising campaigns where the target audience is known to use their ears and eyeballs. In other words, if a film’s target audience is indeed young men (as it frequently is), then a film should be playable among young men, have marketable elements that are thought to appeal to young men, and be advertised through media that young men use. Usually, a film’s target audience is already predetermined 169 when a movie is greenlighted, at which time a full “marketing plan” has already been written. However, a test screening is the first specific empirical investigation into whether or not this plan appears feasible. In the test screening, moviemakers may acquire objective “proof” that everything is going as planned. Or, on the other hand, they may find objective evidence that the film is lacking with its target audience, in which case the film itself may need to be revised—or perhaps even sent directly to video. Although they operate nominally under the guise of “science,” test screenings occur amidst several factors that violate our conventional understanding of science. In contrast to academic sciences, commercial research on movie audiences is explicitly done for money. Not only do audience “research vendors” like OTX, Marketcast, and Nielsen NRG conduct test screenings for profit, but their studio clients also request research for the purpose of making a profit. Timelines for conducting the research are knowingly finite, because a movie has to be released according to a desired schedule. Test screening data has to be readied in time to be used in making decisions by those at the studio marketing the movie. These demands for efficiency have given way to a test screening process that is thus highly routinized. The method for carrying out a test screening does not vary from film to film. Instead, the basic means of audience recruitment, organizing the screening, and administering of audience questionnaires are used repeatedly. One key difference between the commercial audience test screening method and academic science, then, is that the former is probably subject to more explicit efficiency pressures than the latter. Second, audience test screenings are not set up to generate new knowledge, but instead are designed to see how an existing movie project compares within an already 170 existing paradigm. And under no circumstances would data from a movie test screening become a case that is made available for discussion among peers (that is, among fellow audience researchers). Commercial audience researchers are rarely concerned with larger, more circumspect questions about what might constitute an “audience” to begin with. Whereas many academic disciplines value finding new phenomena, introducing new concepts, or revising old concepts, audience test screenings do not lend themselves to producing these outcomes. Instead, they are conceptualized in a more narrow sense, as being useful for simply measuring playability and marketability. Finally, audience test screenings are defined differently than more conventional visions of science because the parties involved in the test screening process are not simply dispassionate observers of an empirical phenomenon (as most academic scientists would at least purport to be). Instead, test screenings are of immense consequence for those connected with the film and their standing in the industry. For studio executives, a test screening may be a cause for either celebration or dread. My instructor from the course on movie marketing described being at test screenings where “the suits” hi-fived each other when they saw that their audience applauded after the title of the test film was initially revealed. 11 On the other hand, if a particular test film is having special difficulty finding audience recruits to sit through a screening (e.g., a low recruit rate), then, as my instructor put, “the marketing department takes out a loaded gun and shoots themselves in the head.” Of course, executives are not the only individuals who have a reputational investment in the outcome of a test screening. Agents, managers, and their acting clients are also are keenly aware that the results of a screening can attribute flaws to a particular 11 How much is revealed to the audience beforehand varies somewhat by who is doing the research. 171 actor’s performance, influence how heavily a film is marketed, and influence how well a film does in theaters—all of which can eventually have a bearing on an actor’s future work. As a source of both potential power and potential dread, the various people connected to a film may be deeply interested in seeing the results of the screening going one way rather than another (a fact whose implications are discussed in more detail in Chapter 5). This too seems to separate audience research from other forms of science. Reproducing Audiences Through Measurement: Test Screenings and Objectivity Problems at the Ground Level The official story that those connected with test screenings might use to describe their work is that “we are doing something objective here.” However, there is a gap between this assumption of objectivity (as an ideal) and how audience researchers actually do their work as well as how test audiences actually behave. These gaps break down along several lines. First, test audiences may not be typical of actual moviegoers, and may therefore misrepresent movie consumers. Second, the theoretical design of the test screening is not completely controlled and is often routinely violated. And third, data collection frequently appears to violate assumptions about audience involvement, leaving the quality of data gleaned from test screenings suspect. Researchers’ (willful) ignorance and even complicity in bringing these problems about indicate, to some extent, that simply producing data may be more important than what that data says. Building a language that appeals in principle to a logic of “objectivity” supercedes fundamental problems with the foundations of that language. 172 Problems Making Test Audiences Representative According to the theory of probability that underlies test screenings, a test screening should be representative of potential moviegoers for that movie. That is, each potential audience member for a particular movie should have an equal chance of taking part in that movie’s test screening. In practice, however, this is impossible for audience researchers to arrange. Afterall, how can anyone actually know who the potential audience is in advance of showing a movie? Nevertheless, decisions about where to conduct a test screening and where to focus audience recruitment must be made, and these decisions are the first steps on the way to formulating technical ideas about who an audience is. Often, these decisions appear to be influenced by cost considerations, taken- for-granted routines, and even “political” considerations between different players within the industry. One of the first selection decisions that must be made is where to conduct the test screening. As Mark, the former studio marketing executive who was my instructor in the course on movie marketing, explained: “How it works is that the directors and executives working on a particular film sit down and decide who they want the audience to be.” Usually, a film has already been cast as either a “blockbuster” movie to be released on thousands of screens or an “independent” film more suitable for release in select theaters (perhaps on only a few hundred screens). The decision about the test screening venue can be an extremely important one for how the film is eventually marketed, because it can reinscribe this distinction between “blockbuster” and “indie,” formalizing it with statistical evidence. Based on the many recruitment announcements I received for different test screenings, such screenings are conducted at a small handful of theaters. 173 Particular companies appeared to have movie theaters that they repeatedly used in their testing operations, with perhaps two theaters that were large megaplexes and one that was a smaller “arthouse” theater. The megaplexes were located in or near large shopping centers that housed various department stores and national chains. The arthouse theaters used in test screenings were always located in areas of town that were at least as affluent as those where the megaplexes were, and were also located among more pedestrian shopping areas that had slightly more upscale shops. Therefore, whether a film was tested at one of the large megaplexes or the arthouse theater seemed to indicate at the outset how the moviemakers conceptualized the audience for the movie, and probably signaled to test audience members that “this film is for me” (thus limiting the potential for a moviegoer who self-identifies as a strictly “independent” movie consumer to come into contact with a film marketed as a more mass-market movie, and vice versa). In other words, the selection of a testing venue helps formalize moviemakers’ conceptions of a target audience before a film is actually tested. As Mark (the former studio executive who taught my business class) also attested, selection of this initial target audience and testing venue sometimes became “political” issues. That is, they could become points of contention for people surrounding a film who had different aims. A director, he said, may have a particular audience in mind for his movie that is at odds with the studio’s idea of the movie’s target audience. He explained how this might play out: The filmmaker has already decided who the audience is, and thinks it will appeal to only, say, men. So the director might want that particular audience because he will want that movie to test well. On the other hand, the studio thinks it might appeal to a different audience or a broader audience, or that screening only to guys would skew the data. Depending 174 on what people’s motivations are, some people—like the director— might want the screening to go very well, even if the screening is not an accurate indicator of how the movie will actually do. This might turn into a big battle. Directors, he informed us, are also particularly prone to raising issues with venue selection. For some, the size of the theater is thought to have an effect on how a film might test, especially for comedies. “It is harder to feel an audience in a large theater” and harder to gauge the crowd’s laughter, Mark told us—which is why some directors of comedies were known to push for smaller theaters. Other preferences for venues seemed to be tied up with individual superstitions, as was the case with a director from the Chicago area who simply preferred to have all of his test screenings done in Chicago. Beyond testing venue and preconceptions of the target audience, there were inherent restrictions on who could be admitted to test screenings. These held true across all of the test screenings I attended. Among these were restrictions based on the appearance of audience members. The recruitment announcements that I received all stated something to the effect that no one would be admitted who appeared “disheveled” or “under the influence” of alcohol or illegal drugs. I speculated that perhaps these statements were made to give researchers leeway to turn away homeless audience members, who appeared absent from test screenings I attended. In addition to this restriction, there were also powerful age restrictions which varied somewhat depending on the movie. Some test screenings required that all audience members be between the ages of 18 and 34. Others allowed for a broader age range, such as 17 to 59. Many of the age restrictions at the younger end of the spectrum were probably related to a movie’s anticipated MPAA rating (in other words, it might create problems to allow someone who 175 is 15 into a movie that will likely be rated R). But given that there are no such restrictions for older moviegoers and that no recruitment announcement I received ever allowed for any test audience member over the age of 60, it appeared as though there were strong assumptions about older moviegoers built into the test screening process. Namely, the assumption was that older moviegoers would not attend this (or any) movie in a substantial way, and, consequently, they should be eliminated from consideration as test subjects and as potential moviegoers. How people are recruited as test audience members may create problems for representing the potential movie audience as well. Test audiences are initially recruited from shopping malls, where they are given flyers that include the time, place, and (sometimes) short plot summary in advance of the test screening. Thus, people who shop more frequently at shopping centers are more likely to be solicited to attend test screenings than those who are less frequent mall shoppers. There are probably some heavy moviegoers who rarely go to malls, although it is unclear the extent to which this results in the misrepresentation of a film’s potential audience. Regardless, I never heard anyone connected with the research end of the business express skepticism about recruiting from shopping malls in principle. Instead, particular problems could arise at particular malls. As, Mark, my movie marketing instructor put it, “sometimes there is a sense that an audience at a particular location has been dried up… The people you start getting are weird.” In these circumstances, Mark said, the testing venue had to be switched up. In addition, problems with “field” were a continual preoccupation for data analysts at the Strategic MediaVision (discussed more below). Here, “field” also involved problems with research sites at malls (see below). 176 The sampling methodology used by test screening companies is what is known in statistics as a “convenience sample.” Test screeners attempted to approximate representation after the fact by setting up demographic quotas. Such recruitment procedures were not geared toward gaining a representative sample as much as they were about making sure an adequate number of people showed up to the screening. This was done largely through soliciting repeat audience members. Once I attended one test screening with a particular company, I was placed on mailing and call lists and was contacted by email and phone recording for each new test screening that fit my demographic profile (that is, a male in his mid-twenties). My wife, who came with me to some test screenings, usually received the same notices as I did, and our cell phones would often ring at precisely the same time with the same automated voice recording soliciting to come out for the same screening. On a few occasions, however, my wife received notices that I did not receive, and was asked to come out for female-only screenings. This strategy of repeatedly contacting a large number of people usually resulted in a very large turnout at the test screening venue. Every test screening I attended completely filled the theater, and on a couple of occasions I arrived too late to get in the theater at all. The extent to which test audiences members were composed of repeat attendees was apparent. At the six test screenings I attended, I had the chance to speak informally with approximately ten of my fellow test audience members regarding their tastes in movies, their experiences at test screenings, and their motivations in coming to them. 12 12 Although none of these individuals ever knew I was doing a research project, this sort of banter was not uncommon among audience members. In some cases, the people I talked with initiated the conversation 177 Most of the people I was able to speak with came to the screening by themselves, rather than in groups. This fact may have skewed the sorts of conversations I was able to have, although there were a large number of people who appeared to be at test screenings alone (perhaps more than I’ve seen in a regular movie theater). Among these individuals, however, it was remarkable to me just how much some audience members reported being involved with not just movie research, but commercial research in general. One man (white, approximately 40 years old, and modestly dressed) that I spoke with told me that he had just been able to make it over to our test screening by bus because he had been participating in a focus group on cell phones on the other side of town. “It was $125 for an hour and a half of work—for talkin’ about nothin’.” He told me that both he and his wife not only went to many movie test screenings, but also many of these paid focus groups—and he also told me how I could get involved if I was interested. While I sat next to him waiting for the test movie to begin, he called his wife on his cell phone and I heard him talking to her about their son. “There’s a cooking show that’s taping, and I was wondering if maybe he’d be old enough to go.” Being a participant in commercial research thus not only appeared to be a significant source of income for this man and his family, but also a significant part of the way his family spent his time. At another screening, I spoke with an African American woman in her late 30’s or early 40’s who reported being involved to a similar extent with commercial research. She told me that she was a screenwriter now and “very hard up,” but that she liked to come to test screenings “to see what’s out there.” Coming to test screenings was a way to and it would have been rude to refuse to make this sort of small talk with people. In all cases, the data I gathered was completely anonymous and only involved several minutes of conversation. 178 work on her craft, and do it on the cheap. “I used to pay for tickets, but now I only come to these.” I did not ask what exactly she did for money, but she told me how she supplemented her income by going to different online market research sites and participating in surveys that paid her in “points.” On one website, she told me that she had earned 4000 points (“That’s like $200!”). When I asked her if it took a long time to complete all of these surveys, she said, “no, you just click whatever” (meaning she just attempted to get through the survey as quickly as possible without regard for what questions actually asked). Although it is not clear exactly how widespread such audience members are at test screenings, all of those who I spoke with that came alone fit this profile (though the above examples are the most striking). Motivations among others I spoke with varied. One group of two white women in their 20’s told me that they came to test screenings because they “liked to have an impact” on what kinds of movies were made. Only one male and female couple in their 20’s told me that they just thought the movie playing sounded interesting, and did not usually come to test screenings. Thus, most of the test audience members I spoke with did not conform to the assumption that test audience members are like conventional moviegoers. They articulated motivations that conventional moviegoers are unlikely to have, such as wanting to have an impact on what films are made. Or, they openly spoke of practices that failed to reflect how conventional moviegoers would attend movies (e.g., by paying for them). Also, their heavy participation in commercial research—whether they did so for an additional source of income or continuous hobby—seemed to make them quite different in their approach to movies than conventional audience members. Compared with the potential American 179 motion picture audience at large, they would perhaps be “outliers” of some sort—what my instructor in movie marketing was speaking of when he talked about the potential for test audiences to become “weird.” Problems Controlling the Test Screening Environment A second difficulty in the carrying out the test screening method, given its assumptions, is the maintenance of a controlled environment. That is, maintaining a test environment where audience members are not influenced by each other or the setting as a whole, where audiences are able to provide feedback that accurately reflects “their views,” and where the study design is executed rigorously. Despite the existence of this ideal, there are several ways in which researchers are unable to control test screenings according to the assumptions of their design. One reality is the potential effect exerted by some segment of the test audience on the perceptions of test audience members as a whole. I experienced an example of how this might happen during one of my regular, paid trips to a local theater in Los Angeles. The movie was David Cronenberg’s A History of Violence, a fairly dark story about a peaceful man who suddenly finds himself reentangled in his violent past. The film, as the title might suggest, contains several gory scenes. Approximately 20 minutes into the movie, perhaps in response to a piece of acting that someone found “cheesy” or perhaps in response to something else, a group in front of me started laughing uncontrollably. Soon other people in the audience became restless and began whispering to themselves, as did my wife, her brother, and myself. We (and others in the theater) had failed to stay immersed in the fantasy of the movie, and we walked home talking about how we 180 thought it was a letdown. However, when we got home and looked up what “the critics” thought of the film, the reviews were generally very good. Other movie aficionados to whom I recounted this story over the course of my research also found it somewhat odd that people would laugh at a movie like A History of Violence, and most people actually told me that they liked the movie. I was left wondering the extent to which I and others in the theater had been influenced by the outburst of laughter a couple of rows ahead of me. Had I seen the movie in a different theater, I wondered, would my companions and I have left talking about the movie in different terms? 13 Statisticians might call this “random error,” and it is an ever-present possibility in audience test screenings. Survey researchers attempt to reduce this by recruiting a larger number of individuals to participate in a survey. If a survey were to consist of ten people, it is quite likely that the opinions collected would not represent the opinions of the population of the whole. Ideally, then, if two surveys were done using a large enough sample in each, their results would be similar. In a similar way, it would be ideal if any two test screenings for the same movie would yield a very close set of average responses to various measures. Yet, movie audiences constitute a social context (Srinivas 2003). Each audience may have its own social-psychological dynamic, making it difficult to know whether a particular dynamic between test audience members was a fluke. In other words, although audience researchers recruit a large enough number of individuals to make their tests representative of the population as a whole (at least in theory), they do not conduct a large enough number of test screenings to know if the social-psychological 13 The same sort of phenomenon is thought to occur in televised politics. Audience members at a Presidential debate, for example, may applaud after a comment made by their candidate. The applauding audience members may only represent the sentiments of an unrepresentative few people, but may nevertheless influence the perceptions of those watching the debate at home. 181 dynamic in a particular test audience is representative. Most movies may only be tested once, and perhaps a few get tested two or three times (often in response to perceived problems with audience responses in earlier screenings). The fact that test audiences (like all audiences) are contexts with their own social psychological dynamics highlights another problem that audience researchers have in controlling the test screening as a setting. Since test screenings are supposed to reflect how actual paying moviegoers would respond to a film, test screenings should not be much different from regular trips to the movies. However, from my first time observing a test screening, I recognized that their emotional tone was different than other trips to movie theaters I had taken throughout my life—in myriad ways. Overall, I see audience test screenings as “event” atmospheres in which audiences are systematically treated differently than they would be if they were to see movies in conventional settings. Attending audience test screenings can feel exciting. One might have the sense that he or she is being let in on a behind-the-scenes world, a sensation that comes from more than just being “one of the first audiences to see a movie.” Test screenings may sometimes take place on studio lots, worlds that are otherwise closed to the public and where actual film and TV production takes place. One recruitment solicitation I received for a test screening spoke of an undisclosed “off-site” location where a movie would be tested. While I could not attend, those who did were instructed to show up at a local theater where they would be bussed to another location. In addition, it is often rumored that celebrities starring in particular movies will be in the test audience. As my movie marketing instructor told our class, “You’ll sometimes see actors wearing ball caps sneaking in the back under the cover of darkness.” Several audience members I spoke 182 with informally reported having seen recognizable actors or directors at screenings they had attended. Finally, routine ways in which the audience test screening is set up may have the effect of “priming” audiences: the act of going through security, having the movie introduced by an enthusiastic master of ceremonies, and the presence of all the research workers and the suits. All of these elements make test screenings feel exciting, like one is in for something unique and special. Experimental researchers have long known that such conditions create a likely environment for those being tested to behave differently than they otherwise would have. A further difficulty in controlling the test screening environment is getting audience research workers to conduct the test screening as planned, according to the rules that are set in place for the screenings to be done in a rigorous way. Workers who carry out audience research sometimes appear complicit in data collection that many researchers might consider “sloppy.” As one example, people are sometimes admitted to screenings who audience researchers did not intend to admit. When audiences are invited to a test screening, they are often given a code that must be used to gain admission. However, at the screenings I attended, I routinely saw people being admitted who did not have the proper code (meaning that they may not have been in the initially targeted group). In such cases, workers carrying out the screening might tell audience members the code, “in case anyone asks.” In a general way as well, I also observed test audience workers sometimes bending the age requirements for admissions to a particular screening. At one screening, the upper age limit for admission to the film was 34. When a man next to me turned in an initial information card that stating that he was 35, the audience worker told him to change his response and simply always say that he was 34. 183 In another instance, a student in my business course on movie marketing also described a similar experience during a class discussion on test screening. We had been talking about the recruitment of test audiences, and the student posed the question: “What about when people make up data?” He recalled a time when he had been recruited to a test screening and the people doing the recruiting had asked him to change his age and information about what movies he had previously seen so that he could take part in the screening. Mark, my course instructor, appeared noticeably perturbed by this question, although he did not deny that this did go on. He responded, a little impatiently, “The system has its fallacies… Yes, there can be negligence. But can there be gross negligence and the studio still stay in business? No.” Similar sorts of issues arose in the recruitment of focus group participants. Although I never made the final cut allowing me to participate in one, these focus groups were conducted following each of the movies that I attended. At a series of test screenings conducted by one research company, audience members were asked to participate in the focus group while standing in line waiting to be admitted. One woman, who looked like she was relatively high up in the test screening operation (given that she sometimes gave other workers directions), would walk up and down the line, scanning the crowd with a searching gaze, as though she were looking for someone she recognized. When she had decided on who she thought might be a suitable target, she approached and would say: “Hi, do you like movies? Do you like talking about movies?” After watching her selection patterns over several screenings, it appeared as though she was selecting on the basis of both demographic criteria (e.g., age, sex, and perhaps race) as well as her perceptions of who she thought would perform well in the focus group. Similar to other 184 workers I saw, she was also willing to bend age requirements when audience members she otherwise wanted to participate were within a few years of the cutoff. She did this with one man who was with his female companion: “Alright, they won’t let you in if you’re 30 or older,” she said softly. “So if anyone asks, just say you’re 29.” Her reason for admitting this man when their were many other 29 or under men to choose from might have had something to do with her second criterion for selecting focus group participants, which was how she perceived that a prospective participant would perform in a focus group. As social scientists involved with focus group research have pointed out, a key assumption of the focus group method is that people actually have something to say. When participants have nothing to say or do not want to speak, it can pose significant problems for this methodology (Morgan 1988). Thus, the recruiter I observed always seemed to be sizing people up, and was sure to tell them that they needed to talk. As I heard her say to a potential participant in a forceful tone, “You’re gonna talk, right? Talk, talk, talk!” Problems with Audience Engagement in Test Screenings The final fissure in data collection at test screenings involves doubts surrounding audience effort in providing information. Dallas Smythe (2001) rightly describes audiences as involved in a kind of “work” for producers. In the case of the test screening, there is an exchange between producers and audiences: producers ask for audience cooperation in providing them with information about audience tastes, and audiences get to see a movie for free. However, this is an exchange relationship in which test audiences are asked to stand in line for long periods and then provide responses to questionnaires, 185 and it sets up a situation in which audience members may attempt to minimize their effort. When this happens, as it habitually does, a key assumption of the test screening method is not met. That is, test audience members may not be providing information that accurately reflects their individual tastes (as provided by a questionnaire). When a test movie ends and the test audience is given questionnaires regarding the movie, many audience members fill out their questionnaires very rapidly—much quicker than they could have possibly read them (much less have considered the answer choices). In watching test screenings repeatedly, a significant number of audience members complete their double-sided questionnaires containing bubble and fill-in responses within 30 seconds. Within two minutes of a test screening ending, over half of the audience is typically gone. Many test audience members, then, attempt to simply flee the theater as quickly as possible. Some test subjects (no doubt accustomed to the process) manage to sneak out right before the credits begin to roll, thereby avoiding any questionnaire whatsoever. Others I observed may simply fill in bubbles as fast as their hand can write. Others still, attempt to leave the movie with partially completed questionnaires. For example, on one occasion, I witnessed a group of teenage boys running down the hall leaving the theater when a test audience worker was able to shout after them, getting them to stop and turn around: “Hey, would guys mind filling out these last few questions?” Those conducting the test screening seem habituated to these aspects of the testing process, and audience workers invariably descend upon test audiences with their questionnaires precisely when a test film ends. As problematic as this reality of test screening may be, some groups of test moviegoers do stay in the theater longer to work on their questionnaires. Not all of these 186 individuals, however, appear to be immersed in thinking about how to respond to the questions. Some groups of test audiences that I observed spent their time filling out their questionnaires together, chatting about their answers and changing them in response to those of their companions. While I sat diligently filling out my questionnaire after each test movie, I was a clear minority in this regard. Overall, then, the test screening process is one that proceeds with significant gaps between the theoretical design of the test screening and its real-world implementation. That is, as practiced, the test screenings I observed routinely failed to meet their own assumptions for pure objectivity. Indeed, real test screenings often seem to face the near impossibility of ever meeting these theoretical assumptions. This not only seemed clear to me as an outside observer, but some of these deficiencies also seemed apparent to many of those involved with the test screening process (even those relatively high up in the research apparatus). This acceptance of audience research as an imperfect process was a pattern I found elsewhere as well, including Strategic MediaVision (where audience data was interpreted) and in my interviews with industry professionals who swam in the world of “objective” audience data. Interpreting Test Screenings: Reifying Audience Measures Once the test audience has completely left the theater and the questionnaires are collected, all that remains are questionnaires with pencil marks, which will then be entered into computers for tabulation and eventual analysis. At this point, problems inherent in collecting data from a real test audience fade into the background as the questionnaire information is now safe for abstraction. The research company or research 187 wing of a studio now kicks into full gear to compile a report which will contain information relevant for the movie’s future. Assuming a movie might not need any more revisions, two questions will be of primary importance at this juncture. First, how is the film predicted to do at the box office? This question might be considered in different ways, based on different mathematical models. In addition to how a movie tests during its test screening, other indicators may be used in formulating this prediction, such as telephone polling that asks audiences about their awareness of upcoming films and their intent to see them. This initial box office estimate is very important for the business of Hollywood films, because it in turn serves as a predictor of how the film is thought to do across other, succeeding venues and formats, including licenses to airlines and cruise ships, DVD sales and rentals, cable television, and network television (what are known in the industry as different “windows of exploitation”). Prediction of box office success based on test screening data is, according to various accounts, sketchy. Although major research firms recently claimed that their projections for a movie’s opening weekend come within 15% of the actual opening about 70% of the time, there are notable examples in which such predictions are profoundly off (Snyder 2006: 49). As Gil Cohen put it during a class session in which he recounted some of his experiences with test screenings, “movies that shouldn’t have worked, worked.” A key informant at Strategic MediaVision told me that this high-stakes game of predicting movies was precisely why she was glad that her company was not involved in test screenings, which she once described as “a messy business.” 14 14 Predicting a film’s success before the movie is even made (what is known as “concept testing”) is known to be even more fraught with peril. Major movie “franchises” such as Harry Potter and Lord of the Rings, for example, are known to have tested very poorly in concept tests. 188 The second question that those involved with a film face as they mull over test screening results deals with what can be done to effectively market the movie. This question turns fundamentally around the notion of the “target audience.” Data gleaned from a test screening speaks to ideas about what groups a film appeals to, and what groups the film fails to appeal to. Using measures provided in the test screening allows studio executives and marketers to “correct” parts of the film or marketing campaign to (hopefully) better appeal to particular segments. Implicit in many of these corrective measures are ideas that certain films should be appealing to particular segments, even though they are not. In a broader sense, however, the measures used in test screenings as well as their routine interpretations by industry professionals occupy a key moment in the construction of Hollywood’s technical knowledge about its audiences. Importantly, I suggest that the data provided in test screenings is gathered and analyzed according to a mechanistic model of audience preferences that limits how audiences might be understood. From both my observations at test screenings as well as at the course I took on movie marketing, I found that questionnaires used in test screenings follow a fairly standard format (see Figure 10 below). The most important measures by far, according to my movie marketing instructor, are the audience’s “overall rating” and their “likelihood to recommend” the movie to friends. Both of these measures are interpreted according to the percentage of the test audience that responds to particular answer choices. For example, the industry accepted “norm” for an audience’s “overall rating” is 55% “positive,” meaning that 55% rated the movie either “Excellent” or “Very good.” Using the logic of test screenings, if 55% of the potential audience for a film thought that a 189 movie was indeed “excellent” or “very good,” one would think that the movie had the makings of a huge hit. In actuality, however, this measure has become somewhat disconnected from what it purports to measure. As my instructor put it, “If you’re at the norm, you’re in trouble.” Ideally, “positive” ratings should be much higher than 55%. Furthermore, my instructor also told us that the proportions of audience members who rated a test movie as “good,” “fair” and “poor” (respectively) did not matter. Fifty-five percent of a test audience could rate a movie as “excellent” or “very good” and 40% could rate the movie as “good,” and the movie would still be considered a probable commercial failure. Responses to the “likelihood to recommend” question are evaluated in a similar way, used as an abstract predictor of a film’s commercial potential. _______________________________________________________________________ Table 7: Some items typically appearing on test screening questionnaires ______________________________________________________________________________________________ Items rated on Likert scales 1. How would you rate this movie overall? (Answer choices: “Excellent” to “Poor”) 2. How likely are you to recommend this movie to a friend? (Answer choices: “Definitely recommend” to “Would not recommend”) 3. How did this movie compare with your expectations? (Answer choices: “Much better than expectations” to “Worse than my expectations”) 4. What do you think of the movie’s pacing? (Answer choices: “Moved too fast,” “About right,” “Moved too slow”) 5. How would you rate the following characters in the movie? (Characters and actors/actresses listed and rated from “Excellent” to “Poor”) Other questions 6. Which of the following describe the film (check all that apply)? (As many as 15-20 descriptors are provided, including miscellaneous possible choices such as “good, Un-PC humor,” “predictable plot,” “a good role for [name of star],” etc.) 7. Where would you expect to see this movie? (Answer choices: “small arthouse theater” or “large megaplex”) 8. Who do you think this movie is primarily for? (Answer choices: “people ages 14-19,” “ages 20-25,” etc.) 9. What were your favorite scenes? (Answer is a fill-in question where respondents write in their answer) ________________________________________________________________________________________________ ________________________________________________________________________________________________ 190 Despite their power in framing interpretations about a movie’s “potential,” there are key uncertainties surrounding what precisely key measures are actually saying. This is especially true in light of the fact that these yardsticks have changed at various moments in the short history of movie research. One interviewee, a movie marketing veteran and former marketing head at a major studio, explained to me with some disdain that several years ago the “norm” for positive ratings was adjusted so that more movies would be seen as hits. In the early 1980’s, only “excellent” ratings were considered to be “positive;” now, both “excellent” and “very good” ratings are counted as “positive.” This adjustment, he explained, occurred because researchers wanted to “please directors,” boosting their egos by being able to tell them that a higher proportion of audiences were seeing their movies as “positive.” Other measures can face similar ambiguities as to how they are interpreted. Mark, my movie marketing instructor, told our class that he “liked” the measure of “whether a movie exceeds your expectations” as an indicator of audience expectations. “If a movie beats the audience’s expectations, that to me is a good sign.” However, he reported that other executives have sometimes turned this notion on its head during discussions of research reports. “They say, ‘why didn’t they [the audience] think it would be good to start with?’” It is not what these measures themselves say that makes them meaningful to industry decision-makers. The ratings of “Excellent” or “Very Good” only have meaning insofar as they are thought to be correlates to box office success. A key epistemological assumption of this mechanistic model, then, is that audiences have measurable preferences that are reducible to useful quantitative scores. 191 However, these measures are used less as sources of information about an audience per se than they are about the movie being tested. In research reports, test audiences themselves are not written about in a circumspect way, and only become meaningful to the extent they are related to a movie being tested. Other questions used in test screenings are employed in a similar way—as measures that are correlates or suggestive of an “average,” but otherwise devoid of analytical value about the audience. Above all, such questions are meant to provide “actionable” data, or information that makes obvious some course of action within the industry’s paradigm of how audiences work. If the movie being tested does not receive a 55% “positive” rating, or scores lower than average on whether or not it will be recommended to a friend, or does not meet expectations, then the movie will have to be marketed differently or released differently. If the movie’s pacing is off in its beginning, middle, or end, then it will have to be shortened or lengthened. If a particular character rates relatively poorly, perhaps there will be less of this character in the final cut. Even fill-in questions asking audience members for their favorite scenes follow this mechanistic model. Potentially rich in information beyond their typical use, they will be tabulated and used to determine which scenes are most “popular.” The scenes that are identified as popular ones are then used in the movie’s eventual theatrical trailers, television spots, and other advertising. These measures of “the audience” serve as institutions of sorts. Few, if any, are sure why these exact audience measures are used or where they came from. They are simply used because they are “industry standards.” As Mark put it in a class on movie marketing, “these are the factors that have been found to be the governing basis for how people end up in movie theaters.” Furthermore, these measures are not subjected to tests 192 or scrutiny to assess whether or not they are still worthy measures. This way of doing research is considerably different than other kinds of “scientific” investigation. Most social scientific research in the academy, for example, is concerned in some way with advancing new concepts or new empirical findings. If academic media scholars had the substantial resources of motion picture studios, they might use this funding largesse to attempt to improve their overall theories of the audience. In routine motion picture research, there is no circumspection of this kind and much less emphasis on complexity and nuance. Indeed, test screenings and other forms of research are probably not well- positioned for these sorts of questions. From the perspectives of those conducting research, a workable paradigm for conducting research exists, institutionalized measures for doing this research are employed in carrying it out, and individual movies are simply plugged in. Some in the business see limits in this model of research. As one veteran studio marketing executive said in an interview with some exasperation: Does the industry understand its customer? Not as well as we could or should considering the amount we spend per year marketing our product. It’s a little bit strange that closer communication with filmgoers doesn’t exist… If I were able to do it, I would like to be able to develop direct communication with my end user. I’d like to know that a 15 year old that goes to see my films is someone I can contact by phone or email, and maybe stay in touch with as they go into their 20’s, and kind of probe a little bit about tastes and what they’re looking at and what they’re thinking. Why not do that?! Because it takes time. And studios are very much run in the way of Lucille Ball in that great skit she did with the cakes on the assembly line. All she had to do was put a cake in a box, and then as the assembly line sped up, she couldn’t keep up with it. That’s the way I view the studios: they’re putting cakes in boxes. Mechanistic and short-term in their vision, audience measurements carry with them important implications. Aside from the industry’s own concern with profitability, some scholars have suggested how the simplicity of these measures may also have 193 consequences for what sorts of movies are ultimately made. Justin Wyatt (1994) ties the rise of marketing research, with its emphasis on quantifiable preferences, as favoring the production of “high concept” films. Wyatt describes high concept films as those that lack narrative complexity and have depthless, image-heavy marketable elements. As kinds of postmodern commodities that were born in the late 1970’s, films are considered to be “high concept” if they can be captured in a title (e.g., 40 Year Old Virgin or Pirates of the Caribbean) or as an image on a movie poster. Such movies require an obvious “marketable” hook of the sort that test screenings are thought capable of uncovering. If a movie does not have a limited set scenes that the test audience identifies as its “favorites,” then it becomes less obvious how to market a film. Wyatt’s work thus suggests how the now institutionalized measures employed in audience research have helped remake the logic of the industry’s decision making calculus, framing those decisions according to abstracted, quantifiable yardsticks. Interpreting Test Screenings: Reifying Audience Categories Although they are not intended as studies of movie audiences in general, test screenings do help reproduce an understanding of movie audiences as differentiated by demographic traits. Mirroring typologies and principles borrowed from consumer research (see Applbaum 2004; Turow 1997), test screenings are organized around and analyzed using categories of consumers that are assumed to be meaningful determinants of moviegoing. These categories are built into the design and analysis of test screenings a priori, such that their explanatory importance is taken for granted and their objective existence as categories is not challenged. This is so despite the fact that these categories 194 are not necessarily “natural.” They are instead social constructs that were created over time, as my analysis in Chapter 2 as well as other critical studies of marketing suggest (see Dávila 2001). There are several key demographic categories that are operationalized in test screenings and movie research in general. First among these are age and sex, summarized by what are known throughout the industry as “the quadrants.” The quadrants include males under 25, females under 25, males 25 and over, and females 25 and older. These four segments of the audience are fundamental to how audience appeal is conceptualized within Hollywood’s technical discourses of its audience, and they are a basis for both audience recruitment and interpretation in test screenings. Among industry professionals, movies are often described by their quadrant (e.g., “Movie X is a first quadrant movie”). Executives formulate guesses and projections about which quadrants a movie will “hit” most as they are deciding which movies to “greenlight,” and such projections enter into final decisions about what films to make (in addition to other factors, such as cost). Mark stated how this mattered, when comparing two hypothetical movies: “Say I have a movie that will cost $40 million and will only appeal to really sophisticated old people, versus this other movie that will appeal to dumb teenage guys and only needs $60 million to break even. It’s a no brainer.” Of course, the belief is that young people go to more movies, so ideally, then, movies should hit with quadrants 1 and 2 (respectively, males and females under 25). Paul, a senior executive at a studio, described this as a kind of “law” that had to be followed if a movie was to be successful. In response to my question on why movies failed, he said: 195 Keep in mind that nobody starts out to make a movie that’s not going to work…. But, certain movies, just don’t work. They look for an audience they can’t get because it’s not young enough, it’s not hip enough. And, it’s too old. It has only one audience. A second important distinction is the division between “arthouse” or “indie” audiences and those seen as “blockbuster audiences.” Arthouse moviegoers are often seen as fundamentally different kinds of spectators, drawn to movies that are different from blockbusters in terms of their small budgets and long release patterns. 15 Within the industry, the typical arthouse moviegoer is typified as more educated, more “urban,” older, whiter, and mostly residing in “blue states.” In a course on the movie business, Gil Cohen made this distinction often. In one instance, we were comparing the then recent successes of two movies: Mr. and Mrs. Smith (2005), which was a big budget, wide- release movie starring Brad Pitt and Angelina Jolie, and Sideways (2004), which had a much smaller initial release that grew into a hit over time. Gil Cohen contrasted the films as follows: “The world of Mr. and Mrs. Smith and Sideways is like— Jupiter is a planet and Mercury is a planet. Sideways played very upscale, it played very urban, blue state, in college towns. It did not play everywhere.” In another course, Gil described a case in which a movie his studio was preparing to release surprised everyone. This was a family comedy that the studio had test screened in California. Based on the results, he and his other executives were bracing for a big loss. Instead, when the movie opened, it was a surprise success. “It was a red state movie,” he said. “It played very well in Oklahoma 15 Blockbusters typically have budgets in the tens of millions of dollars and rely on opening weekends for a substantial share of their box office gross, while arthouse movies typically have much smaller budgets and attempt to build an audience over time (see Wasko 2003). Mark once described the predominant strategy for marketing independent movies as one where you “build the demand—cook it.” 196 and Texas. It played very suburban and non-urban… We live in towers in New York and Los Angeles and have no idea what’s going on.” Thus, ideas of the “blockbuster” audience and “indie” audience operate as demographic types in the Hollywood film industry. They are not only ways of dividing up the potential movie audience, but also work as explanatory factors for why movies work. Boiled down to their cores, the key differences between these two audiences (in the eyes of industry professionals) revolve around broad notions of social class, with these audiences being painted in terms of education, region, and cultural sophistication. Despite this, in the history of American movies, there have been sizable historical shifts in the sorts of movies that different socioeconomic groups watch and the strategies that Hollywood has used in releasing these movies (see, for example, Baumann 2007; Balio 1993). “Independent” films, as they are understood today, are themselves only a fairly recent development on the American moviegoing landscape (Levy 1999). The hard distinction of between “blockbuster” and “arthouse” audiences among today’s industry professionals obscures the shifting nature of these categories. In the course of my interviews and observations, the notion that audiences were separated by both their age and their tastes for either blockbuster or independent movies were largely taken-for-granted. However, on a couple of occasions, the studio executives I interviewed articulated the possibility that Hollywood’s predominant understandings of these variables could be missing something—particularly when it came to age distinctions. Charles Nichols, who was serving as a high-ranking executive at a major studio when I interviewed him, reflected on this point. I had asked him whether he thought it would be possible to target a big “commercial” movie at an older audience, and 197 he recalled an experience he had going to see a movie that his studio had made. The film, he told me, had originally been targeted at people under 30, but ended up opening very strongly among people over 45. I literally went to the opening Saturday night in Santa Monica. I was invited to the premiere, but I didn’t go because movies aren’t the same when you see them with an inside crowd. You want to see them with a regular crowd. There were three ladies, grey haired old ladies who had to be in their 80’s with canes smacking people out of the way to get to their seats. There was a bunch of different groups of people who had to be in their 70’s! …And 80’s! In Santa Monica, on the Promenade on a Saturday night… So in answer to your question, who knows? Mark, my movie marketing instructor, also said that the situation Hollywood found itself in with its age-based “quadrants” was a “chicken and egg” question. “Some people say that if we made more movies for older people, they might go to more movies. It’s hard to know without making those movies, and who wants to do a multi-million dollar experiment like that?” (referring to the high cost of making a typical movie). Overall, if there was a possibility that older moviegoers were indeed “underserved,” the possibility that Hollywood might target older moviegoers was spoken about as though it existed in a parallel universe. Normative demographic types like those involving age and the blockbuster/arthouse distinction, once non-existent, are now thoroughly built into Hollywood’s technical knowledge system. They are integral to how movie research is conducted and how movies are eventually marketed, and work as assumptions that are at least somewhat self-fulfilling. When studio decision-makers decide beforehand on a target audience for a movie, organize a test screening around that target audience, and then market the film at that target audience, it is difficult to know with certainty what would have happened otherwise. Questions on test screening 198 questionnaires that ask audience members which age groups a particular film is for and whether they would expect to see it in an arthouse theater or megaplex (see Figure 1, questions 7 and 8) suggest that audiences themselves may actually be involved in this loop as well. What these questions are really asking is who a test audience thinks a movie is for and not who a movie might actually be for. These questions hint at the complex ways in which research-fed marketing may send signals to movie audiences, simultaneously creating and tapping into consumer identities that say: “this movie is for people like me.” In the process, audience data reaffirms the constellation of categories and measures that comprise Hollywood’s technical discourses about its audiences. Interpreting and Reproducing Audience Data: Life Inside Strategic MediaVision Audience test screenings reveal how technical discourses of audience categories and measures are socially constructed in the process of gathering raw audience data. However, these technical discourses about audiences also shape the routine work of audience researchers as they attempt to make sense of audience data and produce authoritative reports that will be read by various players in the industry. These interpretations are made in the real world of audience researchers, amid the interactional and reputational realities that surround them as they do their work. In the sections below, I describe what this sort of work looks like at a commercial audience research firm, which I call “Strategic MediaVision,” where I worked for approximately two months as an intern and participant observer. 16 In my analysis here, I highlight the importance of 16 See the Methodological Appendix for more on my role as a participant observer at MediaVision. 199 both interactional rules and technical discourses about audiences. 17 These two factors, I hold, interact to shape how audience researchers attempt to write authoritative accounts describing movie audiences. These accounts show how larger technical discourses about movie audiences are reproduced in the everyday work of an audience research firm. MediaVision’s Business Strategic MediaVision was founded in the 1970’s, finding its niche in a movie industry that had just begun to fully embrace the techniques of market research. Located in a mid-sized office park in a suburb northeast of downtown Los Angeles, the company specializes entirely in audience research for the entertainment industries centered in Southern California—one of the few that focus on this market. The company has grown to employ approximately 50 people, who take up two cubicle-filled floors in a mirrored, non-descript four story office building. Tyson, a senior manager at the company, would later describe MediaVision as a “value-added consulting firm.” By this, he meant that MediaVision’s services entailed not just gathering raw data on motion picture and television audiences, but interpreting this information in order to make recommendations to motion picture studios regarding their strategies for selling movies. Depending on the study being commissioned, MediaVision’s studio clients might pay several hundred thousand dollars for these services. The bulk of MediaVision’s current business came from the motion picture 17 Theoretically, this argument is informed in part by the “culture in interaction” approach used by Eliasoph & Lichterman (2003). In this approach, interaction is seen as occurring according to one of a finite number of institutions for forming and maintaining groups. These interactional institutions may in turn interact with larger cultural structures (e.g., repertoires, symbolic boundaries, or codes) to produce social life. I am indebted to this general framework for shedding light on how interactional rules are present in the work of audience researchers and how these rules interact with larger technical scripts that are used in describing movie audiences. As an organizational ethnography, my analysis of MediaVision is also in the spirit of Gouldner’s Patterns of Industrial Bureaucracy (1954), which explored how bureaucratic “rules” were used in interactional contexts. 200 industry, and included studies involving analyses of theatrical trailers, DVD purchase behavior, studies of media technology adoption, as well as some studies that focused on particular audience categories. In all of these studies, the overriding goal was always how to help the studio client improve profits. As Dalia, one of the company’s Vice Presidents told me when we were first introduced and she found out I was a graduate student in sociology, “This is the same thing. All this is sociology for profit.” Making a profit at MediaVision could involve a variety of different methodological approaches to audience research. Many of the company’s studies centered around the use of online surveys, while others involved computer-assisted interviews conducted at shopping malls, telephone surveys, focus groups, and “ethnographies.” 18 Only rarely was MediaVision directly involved in collecting audience data (assuming direct contact with audiences was required). In fact, I was only aware of one study in which an associate at MediaVision participated in conducting a focus group. Rather, the company concerned itself with designing studies, formulating various research instruments (e.g., online surveys, focus group protocols, etc.), and then analyzing data once it came in. Actual data collection was “farmed out,” often to far-flung places throughout the United States and Europe where audiences were being targeted for study. 19 Also, in many studies, several methods were used, presumably to get at research questions from different angles. Much of this, I was told on several occasions, depended on what precisely the client 18 The meaning of “ethnography” in marketing research is somewhat different than in sociology or anthropology. At MediaVision, ethnographies entailed going to media consumers’ homes and looking at how consumers might be using preexisting technology. These studies were not only topically different than most ethnographic studies in academic social sciences, but were generally much more deductive as well. 19 A wide majority of studies involved populations in the United States, although some also involved populations in major markets of Western Europe (Great Britain, France, and Germany). From the standpoint of the entertainment industry, this made a great deal of sense. Approximately half of the Hollywood studios’ global revenue comes from “international” markets, of which Western Europe is a significant part. 201 desired to gain from the study as well as whether the study was deemed financially feasible, from both the perspectives of MediaVision and its client. Making Technical Audience Knowledge in Context: MediaVision’s Physical, Temporal, and Interactional Spaces Work at MediaVision was organized around research projects (i.e., studies) that were being conducted simultaneously. Within the company there were several “workgroups,” each of which was involved to varying extents with a few different projects. Workgroups were comprised of about 6-10 audience researchers and operated semi-autonomously, such that employees generally worked with the same group of coworkers and under the same workgroup leader each day. These workgroup leaders in turn coordinated the company’s overall work with each other and the company’s senior management. This post-bureaucratic organization of work fit with a nominal company rhetoric of “flexibility” (see Boltanski & Chiapello 2005 for more on this sort of rhetoric in contemporary capitalism). I saw signs of this rhetoric beginning on my first day, when I underwent orientation and was given a Powerpoint presentation from Kwan, one of the company’s human resources associates. A key mantra at MediaVision, he said (referring to a Powerpoint slide), was to “Never let rules stand in the way of ideas.” He went on to explain that, “Rules exist for a reason and of course we would never want to break them. But at the same time, rules should never stand in the way of innovation, and rules should never exist for their own sake.” This emphasis on creating an “innovative” environment was also extended to how management allowed the work space and the people inhabiting it to be casual. Martha, 202 my middle-aged workgroup leader and one of my immediate supervisors, described MediaVision as “sort of a young singles environment.” Indeed, many workers at the company were young marketing professionals in their 20’s or early 30’s. But her perspective seemed to extend beyond just the age dynamics of MediaVision’s employees to how the company seemed to eschew any office conventions had little direct bearing on the business of doing audience research. In this way, MediaVision partially reminded me of storied workplaces like Google and other high-tech corporations that encourage informality in the interest of productivity. Although MediaVision was perhaps not the purest example of this sort of work culture, it appeared to borrow from it. “Casual dress” was touted by human resources as a “benefit,” and almost everyone came to work in jeans (unless they were giving a presentation to a client). Inside the office, as well, no particular care was taken to maintain any ideal for the office’s physical space, as was done at Limelight (see Chapter 3). Piles of papers were frequently strewn across employees’ office and cubicle floors, food containers sat on desks, and employees sometimes decorated their own cubicles with odd pop cultural artifacts. Many would play their music at an office appropriate level, including two young men across the hall who would play (with some irony) cheesy pop hits from the 80’s (e.g., Spandau Ballet’s “True” or Phil Collins’ “Sussudio”). In short, MediaVision attempted to construct an office environment where there was little to encumber employees beyond the substantial amount data that flowed through the company. And there was indeed a substantial, never-ending quantity of data to be analyzed. Various project deadlines always loomed, and my workgroup was consistently handling about 10 or 12 projects at any one time—always racing against the clock to keep them on 203 schedule. Research protocols had to be checked, tabulations had to be done, and reports had to be compiled for presentation to studios. In this world, efficiency was a at a premium. Using a centralized computerized scheduling program, days were partitioned up into a series of meetings about specific projects and slots of time for working on those projects. Collaboration was aided by the company’s intranet, where all project data was stored on a single, centralized mainframe so that employees could simultaneously work on various elements of a particular project without duplicating any work. All of this made for an environment that many of my coworkers saw as a “workaholic” one—a fact that became apparent to me when I first started at the company. On my first day at MediaVision, I was given a tour of the company’s facilities by a human resources associate who told me about some of the surrounding amenities. The office park was located near other commercial buildings that included a nearby health club, grocery store, and a privately owned store to mail letters and other packages, all of which she touted as great conveniences. “No one here has time to go to the post office,” she told me with a tinge of humor and astonishment. Many employees seemed to live at work, talking of long days in which they sometimes stayed at the company until 10 p.m. One research manager told me how he would work 14 hour days when he first started at the company. “It killed in the early days,” he said. To facilitate these long hours, MediaVision arranged for a few perks, including making dinners from nearby carryout restaurants available at discounted prices. And, of course, employees headed back and forth to the company’s kitchen throughout the day, where coffee was steadily brewed with a restaurant quality coffeemaker. 204 Time and profit pressures formed the background in which audience research was done at MediaVision. Such pressures may have been endemic to this industry as a whole. When I interviewed the CEO of one of MediaVision’s competitors, he explained the heightening pressures that his company faced, which were similar to those of MediaVision. When I asked him how the audience research business had changed over the years, he characterized how studio clients had continually heightened their expectations: Yeah, I think originally what they [the studios] wanted was really fast information. They just wanted really fast information but not necessarily of high quality. Then what they wanted was good analysis and help thinking through problems. Then what they wanted was they wanted all that cheap. These industry pressures filtered down into the everyday organization of work at MediaVision. Despite company rhetoric about “innovation,” the vast majority of work at MediaVision was very routinized. Paralleling various sociological studies of journalism that show how news narratives are produced against the pressures of time and routine (see Klinenberg 2002; Downie & Kaiser 2002; Gans 1979; Tuchman 1978), research “stories” about movie audiences were produced against similar kinds of pressures at MediaVision. Indeed, the company had developed shortcuts for handling the demands of clients and their tight project schedules. Rather than planning each study anew, MediaVision had developed a “best practices guide” which served as the sourcebook for designing new studies. This manual contained detailed information on how to set up studies of varying methodological persuasions, how to develop plans for implementing the research, how to handle project budgets, and how to present findings to clients. 205 Workers drew on a series of templates for this purpose, including predevised research proposals, instrument protocols, and Powerpoint slides for reports. MediaVision’s data collection had grown to heavily emphasize online surveys and computer-assisted interviews at shopping malls, and the vast majority of day-to-day work at the company was concerned with making sure these instruments were functioning properly. This was done by what was called “protocol checking,” where MediaVision’s employees actually went through the various instruments to test whether respondents were being properly led through survey questions and their “contingencies” (i.e., questions that were to be asked based on previous responses). After survey instruments were up and online, and respondents were actually filling them out, computerized tabulations of responses were uploaded and checked to make sure there was nothing amiss with the online protocol. Almost everyone in the company hierarchy was involved with these sorts of activities on a consistent basis. “Everyone here does protocol checking,” Martha told me toward the beginning of my work at MediaVision. “Right up through Dalia [the Vice President].” Relatively little work time, then, dealt with actual analysis of what data meant. In fact, when I was learning what my job entailed, it often seemed that my initial questions about how MediaVision’s researchers did analytical work came off as perplexing. When Martha was initially explaining the online survey process to me, I was wondering about where questions came from and why particular questions were applied in particular studies. At one point in her explanation, I asked her, “How do you know these questions are asking the right things?” By this, I had meant to get at the issue of question choice and why questions were thought to be valid. Instead, Martha read my question as having 206 to do with how it could be known that the instrument was functioning properly, telling me that, “We have four people in-house go through everything” (referring to protocol checking). Dalia, the Vice President, also underscored the very mechanical nature of work, after finding out I was a Ph.D. student: “We’re probably underutilizing you here,” she said. “This probably involves fewer higher order thinking skills than you’re used to.” Many at the company, however, were very highly educated (most had either an MBA or a Ph.D.) and few, insofar as I could tell, appeared to be drastically more utilized for their higher order thinking skills. 20 Highly rationalized, much of the work at MediaVision took on a very mechanical, redundant quality. For most minutes of the day, audience researchers at MediaVision sat at their computers and dealt with data collection issues or filling in the blanks on reports. Modal forms of work-related talk dealt with these concerns as well. Was the online protocol working correctly? Were tabulations adding up plausibly? Within this environment, I observed signs of alienation from the tasks at hand. Soon after I started at the company, Martha and I were discussing protocol checking when she admitted her frustration with the sort of work required at MediaVision. “We do a lot of dumbass things here,” she said. “As a Ph.D. who worked with SPSS, it was hard for me to believe when I first started.” There was also a tinge of cynicism in this reaction against the routines of checking research instruments and compiling reports, all of which took place in a cubicled office environment. I was struck, for example, by the mock motivational 20 I had a basic sense of what other employees in my workgroup were doing on a daily basis. Every morning, members of my workgroup (including me) had a daily meeting in which we communicated our planned task for the day. Although my workgroup included some more high-level researchers, I noticed that their planned work rarely involved much in the way of analytical thinking about audiences. Instead, it invariably involved some protocol checking and reviewing of tabulations. Higher level researchers also dealt with other problems involving budgets and “field operations” (i.e., sites where data collection had been farmed out). 207 poster that one MediaVision employee had posted on his wall. It read: “Act Like What I Do Actually Matters.” In another revealing moment, as well, I heard Jeff and Trevor (both in their 20’s and in my workgroup) talking about playing video games, one of their shared hobbies. I recorded this exchange in my notes: Jeff and Trevor are talking in the cubicle next to me. He is saying he used to be involved heavily with online gaming. Jeff tells Trevor that when he was unemployed he was playing World of Warcraft like 12 hours a day and lost a bunch of weight. Trevor says that yeah, video games can be addicting and that you could be protocol checking and just thinking about what to do on World of Warcraft when you get home. Jeff then quipped, “What do you think I do when I’m working here?” They both laugh, then Jeff says, “I’m kidding… Mostly.” 21 Although employees in my workgroup dealt with this environment with humor, work at MediaVision meant dealing with a continual stream of problems. Deadlines approached rapidly, and instruments failed to work correctly. This could be cause for considerable stress, because it might threaten to disrupt the project’s schedule and strain relations with clients. When there was a serious problem with an online survey not running properly, the entire workgroup might devote its attention to fixing it. For online studies, this might involve a massive protocol checking operation to pinpoint a problem. During one such crisis, with a scheduled deadline looming and the survey instrument malfunctioning incomprehensibly, my workgroup (along with several other workgroups) began protocol checking a massive online survey on DVD use. A dozen and a half employees had spent the morning meticulously trying to find the “glitch” in the protocol design. A large amount of scheduled work had been put on hold to deal with this 21 Tyson, a senior manager at MediaVision, also had a slightly cynical, if more socially reflexive, view of his work at MediaVision. “Marketing is not completely honest,” he told me during an introductory meeting. “In this business we’re not selling people stuff that’s going to give them cancer or is going to kill them. All we try to do is get people to buy as many DVD’s as possible. The worse thing that happens is the person gets the DVD and says ‘this thing is a piece of crap!’ and throws it out. Fortunately for us, not many people bring DVD’s back.” 208 emergency. “We’re all doing dumb ass shit,” Jeff muttered to me, as we were both checking different pathways in the survey and looking for logical inconsistencies. Frustrating as it was, our efforts were not solving the problem. The test numbers we had been running through the survey in the morning were still coming back incorrectly. Martha came by to look at my tabulations in the early afternoon, and, still perplexed, whispered to herself with frustration: “What the fuck is going on?” Still not knowing why the protocol was not working, she took a moment to educate me on how to think about crises such as this. “You never tell your client about stuff like this. There’s legal issues and so forth.” Eventually, someone who was in contact with the offsite website hosting company that serviced MediaVision discovered that the problem was not with our protocol, but in the way they were hosting the data (unbeknownst to us, they were still using an old version of the survey). Seemingly arcane technical details such as this became sources of collective drudgery on a number of occasions. Thus, for audience researchers at MediaVision, the most frequent risks that were faced on a daily basis had to with research instruments and their potential problems. This affected workers in their interactions with each other, as well. Namely, it provided a prime criterion for evaluating the worth of individuals at the company. Audience researchers lived in a universe of technical competencies, and workers at MediaVision attempted to build reputations within the company as technically competent individuals. This was in contrast to talent managers at Limelight (discussed in Chapter 3), who performed “reputation work” by attempting to meet industry-wide conventions for self- presentation (see Zafirau 2008). Whereas talent managers lived in a world in which perceptions of interpersonal attributes and clout defined individual worth, audience 209 researchers at MediaVision lived in a world in which worth was based on perceptions of technical expertise. In the latter case, then, interactional risks stemmed not so much from the potential loss of clout or betrayal within one’s dense business networks, but from entrusting responsibilities to incompetent people. At MediaVision, employees were described with varying degrees of competence. I gradually noticed how people in my workgroup were recommended for particular functions based not just on their experience, but on whether or not they were seen as competent individuals more generally. Could they handle particular tasks? Were they proactive in navigating technical problems with data collection? In my workgroup, Ethan, a lower-level researcher who had been with MediaVision for approximately two years, provided an ongoing example of how an audience worker at MediaVision could acquire a reputation for weak technical competency. Compared to other workers at MediaVision, Ethan had a meek demeanor. He seldom spoke up at workgroup meetings, and he seemed to keep to himself while at the office, never making small talk with others about life outside of the company. These interactional habits might have contributed to how he was regarded by others in the workgroup, but Ethan was also frequently called out for mistakes with instrument checking and lagging progress with his work. Martha and Kurt, my workgroup’s two supervisors, routinely got short with him. Daily coordination meetings sometimes involved interrogations of Ethan and his progress on particular projects. On one occasion, I inadvertently precipitated an incident in which Ethan was taken to task. Ethan and I had both been working on checking an online survey over a couple of days. This was toward the beginning of my time at MediaVision, and I found what appeared to be a major problem with the way tabulations were adding 210 up on the study. Unbeknownst to me, Ethan had previously checked this portion of the study and had totally missed this problem. With the error brought to light, the study’s timeline was suddenly thrown into limbo. After I re-explained the minutiae of how the tabulations appeared to be behaving to Kurt and Ethan, Kurt began asking Ethan about how he had checked the tabulations and his tabulation checking habits more generally. Ethan began to articulate a plan to email a contact at the web hosting company to get ahead of the problem. Kurt interrupted him, giving him an annoyed, instructional directive. Speaking slower than normal and staring Ethan down, as though he might be speaking to a mentally inferior person, Kurt said, “No. This is a serious problem. This is something that necessitates a phone call. You need to pick up the phone and call him now.” After this incident, I apologized to Ethan for causing him this headache. Ethan replied, “No, you should bring up mistakes when you find them.” Overall, then, the context in which research on movie audiences took place at MediaVision had several noteworthy characteristics. Work was very rationalized and mostly non-analytical, involving fairly rigid routines for conducting proper research and writing reports. For audience researchers at the company, being familiar with the inner workings of specific research instruments and routines for dealing with problems was a prime concern. Reputations within the company were based on perceptions of technical expertise. Coupled with the time demands brought on by tight project schedules, these factors provided little possibility for serious reflection about what data meant for each particular audience research project. This fact became crucial as researchers at MediaVision actually produced knowledge about movie audiences in reports provided to motion picture studios. 211 Doubt, Certainty, and Backstage Data Collection at MediaVision Despite attempts to control and manage the collection and presentation of audience knowledge at MediaVision, the company’s research process in practice was not a pure or seamless one. To use Erving Goffman’s (1959) dramaturgical model of social interaction, there was a significant gap between MediaVision’s “backstage” (where data was collected and analyzed within the company) and its “frontstage” (where finished reports were presented to clients). In this progression from back stage data collection and analysis to front stage presentation of findings to clients, there was a movement from expressing doubts about research to presenting it in a completely confident light (see also Hilgartner 2000). In their frontstage presentation of data to clients (in the form of reports), various inconsistencies and contradictions in the data that had arisen during the backstage data gathering process were always ironed out. Final reports were written as seamless and definite statements on audience tastes and habits. Although researchers were most frequently immersed in mechanical tasks related to how well research instruments were functioning, there were implicit doubts in backstage data collection. Different kinds of doubts surrounded each of the methodologies employed by MediaVision’s researchers. First, one ongoing set of problems in the research process involved what was called “the field.” The term “field” referred to various independently operated research sites where computer-assisted interviews were conducted, which were located in shopping malls around the United States. MediaVision had several full-time “field” specialists who coordinated operations with these field sites. Many of the problems that these specialists and others spoke of in 212 meetings echoed MediaVision’s overall mechanistic approach toward data collection. Was adequate “sample” flowing in from these various sites? Was it flowing in fast enough? And, if so, were the interviews being conducted according to the prescribed protocol? In one planning meeting involving my workgroup, a senior manager, and a field specialist, this last issue became an important focal point. Nader, the field specialist, speculated that the study was not being conducted properly at these various sites. By “properly,” Nader and others in the meeting were focusing on the issue of whether the right questions were being asked, according to the study’s prescribed protocol—not about the quality of the interviews, the recruitment process, or any other qualitative dimensions of data collection. We were discussing possible solutions for this, when someone joked that she would go into a research site in a Los Angeles area mall and pretend that she was a “secret shopper.” She went on, “That way, they won’t be on their best behavior.” This underscored a certain disdain that seemed inherent in perceptions of mall sites. There was little control over what went on at these sites, and the best that could be hoped for was that they would cause as few headaches as possible. Even when the mall studies seemed to “work” on a bare operational level, I did hear some express skepticism about the quality of the data gleaned from computer-assisted interviews. As Kurt commented one time when a group of us were talking about a study currently being conducted, “Sometimes you listen to these tapes and you say to yourself, ‘There’s no way in hell this person is paying attention at this point.’” 22 22 Independently of my work at MediaVision, I did go to one mall-based research site which was, to my knowledge, not connected with MediaVision. Though I did not conduct any systematic analysis, it appeared to face many of the difficulties involved in test screening, as when a recruiter encouraged me to lie about my age. Additionally, the person collecting my responses and entering them into the computer seemed very disengaged, and tended to simplify my interview responses to “qualitative” questions. 213 Other doubts surrounded online surveys, which were heavily used by MediaVision. Specifically, researchers sometimes expressed doubts about respondent involvement with these surveys. Online surveys were good at producing numerical data at low cost, but were those taking it actually paying attention to what was being asked by the survey? Such doubts came to the surface at several points in my fieldwork. During my time at the company, we worked with a number of especially long online studies. Many of these had complex, multi-part questions that did not seem to be inherently interesting. During work on one such study on DVD’s, I heard Trevor (a lower level researcher) and Tyson (a more senior manager) talk about this issue. My notes recorded the exchange: During the morning, Tyson came by Trevor’s desk and asked how things were going with the protocol checking. He had sort of asked this in a skeptical way, like Trevor’s response wouldn’t be good. Trevor: “Yeah, it’s a monster study.” Tyson: “That’s what I hear. I wonder if anyone will run through it once it runs right.” [referring to the survey’s completion rate] Trevor: “Probably not. It’s really lengthy. It asks for these rankings, like 1, 2, and 3 [referring to a list of 20 survey items to be given a 1rst, 2 nd or 3 rd ranking by the respondent]. But that assumes you’ve read all the options… But I understand it’s too late [to fundamentally change it]. I guess the client just wanted certain things.” Tyson: “Man… I’m glad I’m not on it.” Trevor later elaborated on this point as he and I were chatting and the study came up. He told me: You know, you don’t have to have a marketing background or a statistics background to understand people <referring to the respondents>. If you just think about the length of it [the study] and all the options you have to read through… You have to be skeptical of the ‘facts’ <makes quoting gesture> that we as marketing researchers have to work with sometimes. 214 Due to the design of MediaVision’s online surveys, the “facts” were sometimes called into question. Like some of my coworkers, I sometimes had difficulty staying focused on many of the protocols. It was difficult to even read some of them, and I speculated that many respondents (like the one I met during a test screening) would “just fill in whatever” on such surveys. Indeed, I saw some evidence of this in the course of my work, where data appeared to be noticeably questionable. In one study of DVD purchases, for example, about 3% of respondents wrote that they had purchased what seemed like huge quantities (over 100) of DVD’s at various large discount retailers like Wal-Mart and Target over the last six months. Much more probably went unnoticed to me, and some bogus responses might have been undetectable to any researcher. Despite doubts expressed during the course of work, the frontstage stories that researchers told about MediaVision’s methods were sometimes more favorable. On one occasion soon after I started at the company (when I was not yet fully familiar with the company or its research methods), I was having lunch with Martha and Kurt and we were talking about the research methods used at MediaVision. Martha, who had been at the company for over a decade, talked about the online surveys. She told me, “At first I was real skeptical of online surveys. I thought, ‘those people will be different than people who aren’t online.’” She now said, however, that she thought Internet based questionnaires were better for the respondent. “You can get up, walk around, and come back to it later.” I later suspected that Martha had been putting a good face on this type research, for she later seemed to go along with implicit doubts about online surveys. Numbers were known to come back looking “a little funky,” which seemed to suggest 215 deeper flaws in the survey itself. As discussed below, frontstage presentation of data to clients concealed any backstage doubts that researchers expressed about research methodologies. Letting these backstage details out might threaten MediaVision’s business and the collective livelihoods of its employees. 23 Producing Authoritative Reports on Audiences: The Structure of Audience Knowledge at MediaVision At the end of the research process, after all the work that went into perfecting research instruments, MediaVision would produce a finished report for the motion picture studio that commissioned a given study. As mentioned above, the process of writing a report was not rethought with each new study. Rather, reports were structured according to several conventions that, ultimately, limited the sorts of conclusions that could be drawn from audience data. First among these was that the report should contain relatively “simple” findings. After all was said and done, MediaVision’s researchers perceived that those who read their reports (studio executives) were not very interested in nuance or complexity. Martha explained during an informal conversation that prior to writing each report, researchers working on a particular study would have a “story meeting,” which often took place before data was completely collected. As the name would suggest, the purpose of such a meeting was to decide what the overall “story” emerging from the study would be. No matter the study, this story had to “arrive at a single point,” Martha explained. “And it should be easy to understand. I tell people who are working on them for the first few 23 For Goffman, this idea that the frontstage is necessarily contradicted by the backstage is a fundamental one (see Goffman 1959; see also Grazian 2004 for an empirical example of this). 216 times that they should pick two or three really compelling pieces of evidence—not more.” The imperative to keep research stories simple perhaps had something to do with how studio executives were perceived, as well. On a couple of occasions, people in my workgroup alluded to how studio executives were not deep readers, and, furthermore, that “a lot of entertainment people don’t understand statistics that well” (as Martha stated). Simplification of findings came through in the visual presentation of MediaVision reports, as well. Full reports were formatted in a series of professionally-designed Powerpoint slides that contained bullet-point summaries of overall findings and sleek graphics that attempted to condense key ideas. One noteworthy technique used convey report findings involved “typing” audiences by creating fictitious, individual audience profiles that attempted to summarize audience traits (see Figure 1). Such profiles were always cast as fictitious, but they had the effect of reducing data that could be quite complex to its most simple, visually catching elements. A second convention held that each report should provide “actionable” findings. According to Martha’s explanation of this term, this meant that each report should “lead with a recommendation of what the client should do.” Thus, MediaVision systematically emphasized those pieces of knowledge that would provide clients with strategic recommendations. Knowledge of a more theoretical nature or that did not appear to lead to immediate applications for studios were less likely to make it into reports. MediaVision itself made this emphasis on “actionability” the basis for its corporate slogan, which was “Making it Matter.” As Tyson explained this to me, the slogan said that statistics, for instance, could say anything, but what MediaVision did was actually make these statistics matter to clients who were interested in seeing the immediate 217 “value-added” from research. Thus, the idea of “actionability” or “making it matter” was something that was a company focus throughout the research process, from project conception to final report. Indeed, I saw this emphasis echoed in various training manuals on the research process. One Powerpoint training chart I saw was particularly telling. It dealt with quantitative reports, and was entitled “Why reports fail.” In one box, the chart listed that the “incorrect” methodology might be used. 24 The chart explained in a sub-box that a methodology could be “incorrect” because it “does not support decision-making.” Thus, findings that might otherwise be interesting could in fact be worthless if they are not seen as useful to the client’s business strategy. A third very important convention for reports demanded that contradictory, inconvenient facts or significant qualifications be expunged from findings. Instead, report conclusions were always worded in firm, general, and definite terms. My work helping to compile reports most often involved working with focus group data, where I routinely saw how data was made to fit study conclusions. In these studies (which could also be a part of larger studies that included survey data), a report summary detailing key findings was written first. It was then my job to go through actual focus group tapes or rough transcripts find supporting evidence for key conclusions. 25 This job was called “pulling quotes.” After pulling quotes for a variety of focus group projects, I discovered several patterns: themes present in original data that were not emphasized in key findings, conclusions made based on scant or nonexistent supporting evidence, and even a few cases in which statements focus group participants made contradicted key “findings.” 24 The word “incorrect” was actually in quotations, which I took to mean that those writing it understood that it might not objectively be the incorrect method, but more a methodology not suited for the report’s actionable purpose. 25 At times, “transcripts” were extremely rough. They usually included many sentence fragments and paraphrasings of respondent data. Only rarely in each document was there an extended, verbatim quote. 218 Figure 11: Sample demographic profile Sample demographic profile obtained from The Intelligence Group, a large market research firm specializing in marketing to tweens, teens, and young adults. The company services major entertainment firms, among other companies. The profile shown here is similar to profiles used in report at MediaVision. (Retrieved from www.youthintelligence.com in 2006) Finding supporting evidence for key findings was my defined task while pulling quotes, so it became my most immediate concern. Midway through doing my first “quote 219 pulling” exercise, I was having a hard time finding enough quotes to support several key conclusions and went to Kurt in his office to ask for guidance, where I had the following exchange (recorded from my notes): I told him I was having a hard time finding enough quotes to support some of the conclusions Also, I told him, there were sometimes very brief notes on what people had said that were hard to use as actual quotes. He laughed sort of an embarrassed snicker and started mock-banging his head on his desk, jokingly. He then started trying to explain, like he was looking for the right words. Trying to fill in the blank, I asked, “Do you want me to just try to make them sound like things people would actually say?” [referring to the notes on what people had said]. He repeated my suggestion, “Yeah, make them sound like things people actually said…,” but he was still sort of smiling, like that wasn’t really it. Finally, he just said straightly, “Make them up.” I was like, “make them up?” And he kept smiling, saying, yeah, just make them up. He then tried to justify it, like it was completely legit, saying “Martha wrote this [the report], so we already know it’s true.” “Besides,” he said with a dismissive shrug, “It’s qualitative.” We both were sort of smiling this knowing smile, and as I left his office he said with what seemed like a little embarrassment, “Sorry.” My second day pulling quotes (for the same study), I was finding a great many quotes that appeared to actually contradict a key finding. The contradictory quotes were so consistent that I was afraid that Martha had made a simple mistake and stated the opposite of what she had intended to write. I returned to Kurt’s office to ask him about it. He appeared sort of frenzied, and at first he thought I was asking him the same question I had asked him the day before. “Make them up,” he said a little impatiently, before I could fully explain my question. When I further explained that not finding enough quotes was no longer my current problem, but it was instead that I was finding contradictory quotes, he seemed to understand and said, “Oh, shoot…” After thinking for a moment, and again smiling as though it were a familiar joke, he instructed me on 220 how to handle these quotes: “Just make them very vague. Like so vague you couldn’t possibly know where it came from.” I left his office not knowing exactly what this meant, especially since his solution did not seem to quite address my difficulty with contradictory evidence. But I finally had the impression that my superiors simply did not want to be bothered with problems “pulling quotes,” so I continued to improvise— sometimes by completely fabricating quotes. 26 Some of my coworkers seemed to find similar difficulties with focus groups. As I was watching a taped focus group one day, Travis (a research associate) came by my desk to chat. Commenting on the job I had before me, he said, in a sort of joking tone, “Oh yeah, it’s rough…” Confirming my own experiences, he also said, “Yeah, it can be hard to find quotes that match what’s in the summary.” He went on, reflecting on how our superiors wrote these summaries, “Sometimes it’s hard to know what they had in mind [when they were writing the summaries].” But, he acknowledged, “it must be hard to tell a story out of the chaos of comments.” When working with focus group reports, I was also instructed to take other shortcuts in order to firm up findings which would eventually be presented to clients. For example, I was told by Martha to “make sure not all the quotes come from one person or one place.” This applied frequently in studies that involved focus groups in multiple cities. In such cases, I was told I could attribute quotes gained from one city to people from other cities to make the responses appear more generalizable. Where applicable, Martha told me also to try to mix up the ethnicities of the respondents, and that it did not 26 Kurt was not the only one of my superiors who seemed aware of this practice. After my second day of working with focus group quotes, Martha came over to my desk and said, “you saw how to do that, right?” She said this with what I interpreted as a knowing look, as if there was some implicit meaning to her question. I was aware of the ethical dilemmas inherent in this sort of work. Yet, fabricating quotations seemed to be so thoroughly a part of MediaVision’s analysis of focus groups that it would have been hard to resist this practice and still remain a participant observer. 221 matter so much exactly who said what. Furthermore, I should make all quotes sound “grammatically correct” and word them in full sentences. Sometimes, the shortcuts were quite blatant. In one study involving international focus group data, for example, I was instructed to use the report summary from the U.S. “as a basis” for different countries in Western Europe—ignoring the possibility that there were significant national differences between respondents. What I gradually discovered in working with focus groups was that senior management frequently went to focus groups and gave them a cursory listen, typed up relatively brief notes, and then wrote the report. There was little deep or systematic analysis of focus group data. “Telling a story out of the chaos of comments,” as Travis had put it, was exactly the problem that confronted audience researchers on a day to day basis. The need to tell a simple, actionable “story” in a very short time span made the task of writing a complex—even accurate—account of audience data almost impossible. Against this backdrop of meeting deadlines and interpersonal worries about appearing “competent,” the work of making stories about audiences could be greatly facilitated by drawing on familiar, overarching narratives about audiences that were known to have worked. Instead of writing new analytical stories for each project, those who wrote summaries appeared instead to focus on efficiently writing plausible ones. As discussed in the next section, these stories frequently drew on familiar technical assumptions about audiences that were widely held elsewhere in the industry. 222 Knowledge Made Within Preexisting Audience Categories and Demographic Types In the course of attempting to formulate authoritative conclusions for studio clients, audience researchers effectively helped reproduce already existing assumptions about audience groups. This was so despite significant gaps between “findings” and the actual data that these findings were based upon. It is difficult to know with certainty just how far “off” these stories were. However, what I did find noteworthy was how the interpretive work of researchers at MediaVision was frequently guided by dominant stories about demographic groups. Data was not allowed to drift far from these storylines, and when researchers were confronted with confounding data, it was mostly greeted with skepticism. In his Structure of Scientific Revolutions, which examined how “paradigm shifts” occur among academic scientists, Thomas Kuhn (1962) referred to this very phenomenon. Findings that challenge dominant paradigms within academic disciplines, he argued, are more often questioned, sidelined, and even ridiculed compared with those that support the dominant paradigm. In the same way, demographic audience categories were part of the audience “paradigm” that researchers worked with at MediaVision, and these categories resisted change in a number of instances that I observed. One such category was that of the “independent” movie consumer. I had a chance to work closely on one study that dealt with the movie viewing and DVD purchase habits of “independent” movie consumers, which involved focus group research from a variety of U.S. cities. During the course of “pulling quotes” for this study, I was able to systematically note the frequent discrepancies between MediaVision’s final report on the study and what participants actually said (as I was responsible for going through all the 223 transcripts and was able to watch much of the original video from these groups). One key claim in the report was that independent moviegoers were a distinct breed who almost exclusively watched independent films and shunned “blockbusters.” In addition, the report stated that independent filmgoers were more politically progressive, valued more thoughtful fare, relished cable channels like Sundance and IFC (both of which feature independent films), and carefully considered film reviews from more prestigious newspapers like the New York Times. This depiction of the independent audience seemed to square with a general picture that I saw presented elsewhere in the industry. Independent moviegoers were thought to be cosmopolitan, be less passive than “blockbuster” audiences in their movie selections, and even a little snobbish in their rejection of all things deemed part of mass culture. What I found is that the actual data that formed the basis for this report did not entirely support this depiction. While there were elements in the data that agreed with this stereotype of “indie” audiences, there were also other strains that seemed to openly contradict this picture—or, at the very least, provide a more complex view. A few focus group participants (a minority) seemed to state in one way or another that they did not like “blockbusters.” They had a clear notion of what they considered a “blockbuster” (most often, a special-effects heavy film), and they did not like it. However, there were a sizable number of focus group participants that seemed to express no particular affinity for either independent movies or blockbusters. One man in one group said that most of his favorite movies could indeed be considered blockbusters, but were still “enthralling.” Another said something similar when he expressed his preference for movies that were “fun” and “not to serious,” citing examples from a number of popular big-budget 224 comedies. Others described how they collected all kinds of DVD’s, not just DVD’s of independent films. Given these statements, many of the people I watched on focus group tapes seemed more like omnivorous movie lovers than snobbish univores who only liked independent films. Some other claims from the report also appeared suspect after closer examination. For example, several respondents did not live up to their portrayals as avid readers of more prestigious publications or watchers of various independent movie channels. Some respondents stated that they got much of their information from far less prestigious publications like TV Guide. Some said they would simply go to theaters and see what was playing. In the same vein, when it came to purchasing DVD’s, some said they had no problem going to decidedly lower-brow stores like Wal-Mart to get their movies. Finally, it should be noted that some resisted the idea of being characterized as consumers at all, which was not made any mention of in reports. 27 Again, it was not that some subjects did not agree with some of the report’s conclusions some of the time. Rather, it was the fact that the report had worded its conclusions in such a way that it sounded like all subjects agreed with all of these statements all of the time. Based on the data, independent moviegoers were perhaps not such a distinct group afterall. I did not see any comparable study dealing with “young” audiences while I was observing at MediaVision. But this audience category still made frequent appearances in researchers’ interpretive work. As was the case with “independent” audiences, the category of “youth” carried with it a meta-storyline that resisted new interpretations, even 27 This was a very frequent occurrence in my examination of focus group data. While watching focus group tapes, I routinely saw how many participants tried to resist being classified as consumers. If participants were not heavy shoppers or indicated complete contentment with the material goods they already had, such statements never made it into final reports. I attributed this, again, to MediaVision’s emphasis on “actionability,” as non-consumption seems to be the least actionable finding possible for market researchers. 225 when data might suggest otherwise. During informal meetings and conversations that I observed, I saw that researchers practiced a kind of selective skepticism when it came to data that would fundamentally question assumptions about “youth” and the kinds of movies they liked. This occurred in the analysis of one DVD title that was initially seen as a “horror” picture, which is a genre that is seen as drawing heavily from the teenage male demographic. However, with this particular title, the data were not showing this. Numbers on interest in the title were coming in surprisingly “old,” and Sameer (a mid- level researcher) and Kurt were talking about this across the hall. I recorded this exchange in my notes: Kurt was saying, “Why are the numbers coming in so old?” Sameer and Kurt continued to brainstorm over it, like they were trying to figure it out. Then Sameer was like, “Have you seen it?” [meaning had Kurt seen the movie]. He paused for a second and then said, “that’s suspense. That is not what young males are interested in. It’s a totally different target.” They kept talking about it for another minute, but still didn’t seem to have made up their minds when Sameer left Kurt’s office. Kurt was asking again, “So, why is it so old? We’re gonna have to have some sort of story to back it up” [speaking of when they would present their data to their client]. On his way out, Sameer was like, “Yeah, I agree.” When Kurt and Sameer were confronted with confounding evidence about “youth,” they had tried to reinterpret data by playing with how they might categorize the film’s genre. The conclusion that preferences for a horror movie might skew old was not one that they were willing to draw. On another occasion, during a workgroup meeting, I saw again how expectations surrounded the “youth” audience and its movie preferences. We had been discussing a movie title that was based on a video game and what might eventually be said about it in the story meeting. Sameer had not seen the data, but said to Martha with confidence: “Yeah, I would expect it to be heavily young men.” This statement 226 showed how preexisting expectations could inform researchers’ interpretations at MediaVision. The presence of these preexisting discourses about demographic groups showed how audience research was structured at MediaVision. Each new project was not a blank slate in which the data naturally produced a logical set of conclusions. Rather, at MediaVision, interpretation of audience data was situated among a number of constraints: a rationalized, mechanical research process that was almost exclusively focused on the operability of research instruments; an interactional world where researchers attempted to maintain a reputable, technically competent front; and conventions for writing reports that demanded an authoritative tone and limited the sorts of conclusions that could be drawn. Amid these pressures, technical audience discourses—grounded in the idea of distinct demographic categories—themselves became conventions. They were convenient for researchers who worked to come up with plausible stories to present to their clients. And, if they were convenient, they were also feared, because to attempt to reformulate how audience categories like “independent” and “youth” were defined would have been potentially very risky. Conclusion In various examinations of audience research, there are two familiar perspectives on how media producers utilize scientific methods to assess audience tastes. One, stemming largely from economists, marketers, and media professionals themselves holds that audiences are objectively measured. Accordingly, audience groups and their tastes can be identified using systematic methods. Measurement may be error prone, but the 227 techniques used are generally robust. In this way, audience research plays a vital role in helping to address informational problems that would otherwise plague an industry like Hollywood. People, in other words, are more or less getting “what they want.” A second, very different perspective offers a much more critical assessment. Various media critics have attempted to fight against the perceived power of marketing- driven audience research. Many writing from this vantage point see marketing as part of an evil media system capable of “getting inside our heads” in seductive, even subliminal ways. In this light, people are also getting what they “want,” only they don’t really want it because they are being tricked. What I hope the foregoing analysis suggests is that perhaps both of these perspectives might be giving audience researchers too much credit. The audience research that I observed during my fieldwork reveals a flawed, at times hapless, process. Out of this process, movie consumers are made as much as they are measured. Or, more precisely, technical ideas of audiences are constructed through measurement itself. The fact that dominant stories about audiences are able to withstand contrary and indeterminate evidence is testament to how audience categories have been reified as “objective” facts. In the process, an entire technical language for speaking about movie audiences has come into existence, and has proven quite durable. One noteworthy feature of movie research is the extent to which researchers themselves are complicit in reperforming a “science” that they know to be flawed. Doubts about audience research extend beyond the walls of firms like MediaVision. Despite attempts to firm up technical knowledge by researchers, this form of knowledge is unevenly adopted by others in the industry—including the studio executives who make 228 ultimate decisions about movies. However institutionalized this technical knowledge is as a language, it has not led to definite patterns in the ways studios actually make decisions. This partial decoupling of technical audience knowledge from how it is acted upon is discussed in the next chapter, as I show how both intuitive and technical audience discourses work as “justifications” when decisions about Hollywood movies are made. 229 Chapter 5: The Hollywood Film Industry, Evaluation in Interaction, and Organizational Coordination Using the comparative cases of Limelight Entertainment and Strategic MediaVision, this study has examined two factors in organizational coordination: shared conventions for evaluating products and decisions, and the shared conventions for interaction that stem from organizational conditions. Both of these cases illustrate the close connection between these two factors, but should also speak to the extensive variation that is possible in organizational coordination. It is important to note that there are many possible ways in which organizational discourses of evaluation might interact with conventions for evalaution—namely, because there are many different ways in which orders of worth might be configured, a number of possible interactional conventions within organizations, and therefore a great many specific ways in which “evaluation in interaction” might be configured. The evaluation in interaction approach I have offered here suggests greater attention to these complexities. I see this approach as useful within the sociology of organizations, but I also see it as potentially useful in studies of mass media production. This study represents a departure from other paradigms for understanding cultural production within large-scale organizations that produce popular culture. First, it joins a list of sociological rejoinders to studies of the media emanating from neoclassical economics, which tend to begin with the notion that cultural producers are rational profit maximizers and work out from there. My focus in this study is also substantially different from sociological approaches cultural production that see the mass media as a manifestation of power (to the extent that “power” is conceived in terms of one group with essential interests and/or dispositions 230 dominating another). Marxist approaches, for example, tend to see cultural producers as either conscious manipulators or as extensions of hegemonic processes that ultimately serve to reproduce class domination. In a somewhat similar vein, Bourdieu’s sociology understands mass media production as grounded in the class-based “habituses” of cultural producers, which ultimately works to continually reconstruct “the field of power” and reproduce social inequality. Although I see these sorts of analyses as valuable, my approach in this study has examined mass media production using a different definition of “power.” Here, I have dealt with various, interacting constraints that ultimately help shape the actions and talk of cultural producers, regardless of their positions in larger social hierarchies. Given this focus, the evalaution in interaction approach draws attention back to media production as an organizational phenomenon. In this sense, this approach shares commonality with the “production of culture” perspective advocated by Peterson and others (see Peterson & Anand 2004), although my approach places less emphasis on the structural dimensions of cultural production. Furthermore, I attempt to more specifically theorize media organizations as evaluating organizations. Herein, I see organizations like Limelight and MediaVision as constituted by multiple, interacting cultural structures. Collectively, these cultural structures form a powerful (if contingent) basis for organizational coordination in Hollywood moviemaking. Potential variations within the evaluation in interaction approach At the outset of this study, I defined the evalaution in interaction approach as one that was concerned with organizational discourses of evaluation as historically formed, 231 interactionally contingent modes of coordination. These “institutionalized discourses” (which, in this study, included intuitive and technical ways of speaking) had four important features: they were durable, they were (specifically) modes of evaluation, they had an epistemic quality, and they were deployed within interactional conventions. Using the evaluation in interaction approach, these features represent possible ways to think about how organizational coordination varies. The first feature of the institutionalized discourses described here (durability) assumes the existence and salience of institutionalized scripts for action within organization. Neoinstitutional theorists as well as Boltanski & Thévenot both see organizational life as coordinated by a finite number of regular, durable, and transposable scripts for talk and action. However, an important caveat is that such scripts may not be active at all times, even in organizations that are routinely governed by these practices. For Boltanski & Thévenot, there are important exceptions when discourses evaluation cease to be relevant (i.e., when worth is justified purely on individual emotional grounds, or when decisions are kept private). Thus, the evaluation in interaction approach cannot describe how evaluating organizations work all of the time, but perhaps only how they may work most of the time. This caveat aside, there are a number of lines along which organizations engaging in evaluation might vary. These include variations in which orders of worth are operating within a particular organization or organizational field. In the business of making Hollywood movies, I have argued that these primarily involve “intuitive” and “technical” justifications. But other orders of worth might emphasize different principles of worth, such as “creativity,” “fame,” “harmoniousness,” or “environmental friendliness.” As 232 Boltanski & Thévenot point out, these orders of worth can (and do) combine in any number of ways within the contemporary “composite” organization. Another line of variation has to with how institutionalized discourses of evaluation function as epistemic languages. In one sense, these discourses answer the question of “how we know what has worth.” But, in a more specific sense, they involve constellations of terms and objects whose meaning and reality has been constructed over time. Thus, within particular orders of worth, the particular concepts might be profoundly different had their history been different. In the case of Hollywood, for example, the idea of a “commercial” film might not have ever been associated with young audiences had the industry’s historical circumstances been different. Thus, an important aspect of the evaluation in interaction approach is to be attentive to specific concepts that exist as meaningful objects within more general “orders of worth.” Finally, institutions for engaging in legitimate interaction can vary, depending on organizational circumstances. Some interactional institutions (like those described in the cases of Limelight and MediaVision) define legitimate “groupness” as doing various kinds of reputation building. Still other interactional institutions (such as those found in civic organizations) might be less focused on “reputation work,” and more focused on the idea of having deliberative discussions. These different styles of doing groupness can help set the stage for which kinds of evalaution are used, and to what extent. Taken together, these possible variations demand attention to the specific contours and contexts of evaluation within particular organizations. As an approach to organizational analysis, this project involves defining evaluation as multi-faceted discourses that work as a means for organizational coordination. And, it involves 233 locating these discourses within the particular interactional conventions in which they are deployed. 234 Bibliography Abel, Richard. 2006. Americanizing the Movies and “Movie-Mad” Audiences, 1910- 1914. Berkeley and Los Angeles: University of California Press. Acland, Charles R. 2003. Screen Traffic: Movies, Multiplexes, and Global Culture. Durham, North Carolina: Duke University Press. Ahlkvist, J. A. and R. Faulkner. 2002. “ ‘Will This Record Work for Us?”: Managing Music Formats in Commercial Radio.” Qualitative Sociology 25(2): 189-215. Ang, Ien. 1991. Desperately Seeking the Audience. New York and London: Routledge. Applbaum, Kalman. 2004. The Marketing Era: From Professional Practice to Global Provisioning. New York and London: Routledge. Armstrong, Elizabeth A. 2002. Forging Gay Identities: Organizing Sexuality in San Francisco, 1950-1994. Chicago: University of Chicago Press. Baker, Wayne E. and Robert R. Faulkner. 1991. “Role as a Resource in the Hollywood Film Industry.” American Journal of Sociology 97: 279-309. Balio, Tino. 1993. Grand Design: Hollywood as a Modern Business Enterprise, 1930- 1939. History of the American Cinema, Vol. 5. Berkeley and Los Angeles: University of California Press. Banet-Weiser, Sarah. 2007. Kids Rule! Nickelodeon and Consumer Citizenship. Durham and London: Duke University Press. “Banished Leo the Lion is Back; Old Trademark Symbol of Split, ‘Traditional’ & ‘Now’ Show Biz.” 1970. Variety, May 20. Baumann, Shyon. 2007. Hollywood Highbrow: From Entertainment to Art. Princeton: Princeton University Press. Becker, Penny Edgel. 1999. Congregations in Conflict: Cultural Models of Local Religious Life. Cambridge: Cambridge University Press. Benson, Rodney and Erik Neveu, eds. 2005. Bourdieu and the Journalistic Field. Malden, MA: Polity. Bielby, Denise D. and William T. Bielby. 1999. “Organizational Mediation of Project- Based Labor Markets: Talent Agencies and the Careers of Screenwriters.” American Sociological Review 64: 64-85. 235 -----. 1996. “Women and Men in Film: Gender Inequality Among Writers in a Culture Industry.” Gender & Society 10: 248-270. -----. 1994. “‘All Hits Are Flukes:’ Institutionalized Decision Making and the Rhetoric of Network Prime-Time Development.” American Journal of Sociology 99(5): 1287-1313. “Big Coin Aud Still ‘Family’: Avoids Gastritis in the Boondocks.” 1970. Variety, April 3. “Bigger Xmas-New Year Boxoffice.” Variety, January 14. Binder, Amy. 2002. Contentious Curricula: Afrocentrism and Creationism in American Public Schools. Princeton: Princeton University Press. Biskind, Peter. 1998. Easy Riders, Raging Bulls: How the Sex-Drugs-And Rock 'N Roll Generation Saved Hollywood. New York: Simon & Schuster. Boltanski, Luc and Eve Chiapello. 2005. The New Spirit of Capitalism. Trans. Gregory Elliott. London and New York: Verso. Boltanski, Luc and Laurent Thévenot. 2006. On Justification: Economies of Worth. Trans. Catherine Porter. Princeton and Oxford: Princeton University Press. -----. 1999. “The Sociology of Critical Capacity.” European Journal of Social Theory 2: 359-377. Bourdieu, Pierre. 1993. The Field of Cultural Production: Essays on Art and Literature. Randal Johnson, ed. New York: Columbia University Press. -----. 1977. Outline of a Theory of Practice. Cambridge: Cambridge University Press. Bruno, Michael. 1970. “N.Y.’s Gadabout Film Buffs.” Variety, April 29. Burawoy, Michael. 1999. “The Extended Case Method.” Sociological Theory 16(1): 4- 33. “‘Can’t Let Youthful Buffs Hog Screen Fare,’ Sez Seaton.” 1970. Variety, April 1. Cantor, Muriel G. 1971. The Hollywood TV Producer: His Work and His Audience. New York: Basic Books. Cantor, Muriel G. and Elizabeth Jones. 1983. “Creating Fiction for Women.” Communication Research 10: 111-137. 236 Castells, Manuel. 2000. The Rise of Network Society. 2 nd ed. Malden, Massachussetts: Blackwell. Cohen, Stanley. 1972. Folk Devils and Moral Panics, London: MacGibb and Kee. Clemens, Elisabeth. 1993. "Organizational Repertoires and Institutional Change: Women's Groups and the Transformation of U.S. Politics, 1890-1920." American Journal of Sociology 98: 755-98. Cruz, Jon. 1999. Culture on the Margins: The Black Spiritual and the Rise of American Cultural Interpretation. Princeton: Princeton University Press. Dávila, Arlene. 2001. Latinos, Inc.: The Marketing and Making of a People. Berkeley and Los Angeles: University of California Press. De Vany, Arthur and David Walls. 2002. “Does Hollywood Make Too Many R-rated Movies? Risk, Stochastic Dominance, and the Illusion of Expectation.” The Journal of Business 75: 425–51. Desrosières, Alain. 1998. The Politics of Large Numbers: A History of Statistical Reasoning. Trans. Camille Naish. Cambridge and London: Harvard University Press. DiMaggio, Paul J. and Walter W. Powell. 1991. "Introduction." In The New Institutionalism in Organizational Analysis, Pp. 1-38. Walter W. Powell and Paul J. DiMaggio, eds. Chicago and London: University of Chicago Press. “Drive to ‘Poll Your Own Theatre’ To Find Out Why Patrons Stay Away.” 1950. Variety, May 3. Downie, Leonard Jr. and Robert G. Kaiser. 2002. The News About the News: American Journalism in Peril. New York: Alfred A. Knopf. Durkheim, Emile. 1961 [1912]. The Elementary Forms of the Religious Life. New York: Collier Books. Edelman, Lauren B. 1992. “Legal Ambiguity and Symbolic Structures: Organizational Mediation of Civil Rights Law.” American Journal of Sociology 97: 1531-1576. Eliasoph, Nina and Paul Lichterman. 2003. "Culture in Interaction." American Journal of Sociology 108(4): 735-94. Espinosa, Paul. 1982. “The Audience In the Text: Ethnographic Observations of a Hollywood Story Conference.” Media, Culture & Society 4: 77-86. 237 Ettema, James S. and D. Charles Whitney. 1994. Audiencemaking: How the Media Create the Audience. Thousand Oaks and London: Sage. Foucault, Michel. 1980. Power/Knowledge: Selected Interviews and Other Writings, 1972-1977. New York: Pantheon Books. -----. 1978. The History of Sexuality. Vol. 1. Trans. Robert Hurley. New York: Vintage. -----. 1977. Discipline and Punish: The Birth of the Prison. Trans. Alan Sheridan. New York: Pantheon Books. -----. 1972. The Archaeology of Knowledge. New York: Routledge. Friend, Tad. 2005. “Secret Agent Man.” The New Yorker, 21 Mar 2005, 48-61. Gaines, Donna. 1991. Teenage Wasteland: Suburbia’s Dead End Kids. Chicago: University of Chicago Press. Galison, Peter. 1997. Image and Logic: A Material Culture of Microphysics. Chicago and London: University of Chicago Press. Gamson, Joshua. 1998. Freaks Talk Back: Tabloid Talk Shows and Sexual Nonconformity. Chicago and London: University of Chicago Press. -----. 1994. Claims to Fame: Celebrity in Contemporary America. Berkeley and Los Angeles: University of California Press. Gans, Herbert J. 1979. Deciding What's News: A Study of CBS Evening News, NBC Nightly News, Newsweek, and Time. New York: Vintage. Geertz, Clifford. 1973. "Thick Description: Toward an Interpretive Theory of Culture." In The Interpretation of Cultures: Selected Essays, Pp. 3-20. New York: Basic Books. Gergen, Kenneth J. 1991. The Saturated Self: Dilemmas of Identity in Contemporary Life. New York: Basic Books. Giddens, Anthony. 1991. Modernity and Self-identity: Self and Society in the Late Modern Age. Stanford: Stanford University Press. Gieryn, Thomas. 1983. "Boundary-Work and the Demarcation of Science from Non- Science: Strains and Interests in Professional Ideologies of Scientists." American Sociological Review 48: 781-95. 238 Gitlin, Todd. 1983. Inside Prime Time. Berkeley and Los Angeles: University of California Press. Gladwell, Malcolm. 2006. “The Formula: What If You Built a Machine to Predict Hit Movies?” The New Yorker, Oct. 16. Glaser, Barney and Anselm Strauss. 1967. The Discovery of Grounded Theory. Chicago: Aldine. Goffman, Erving. 1977. “The Arrangement Between the Sexes.” Theory & Society, 4, 301-331. -----. 1967. Interaction Ritual: Essays on Face-to-Face Interaction. Garden City, New York: Doubleday & Company. -----. 1959. The Presentation of Self in Everyday Life. Garden City, New York: Doubleday & Company. Gold, Richard. 1986. “Comedy Tops Summer Yoksoffice: ‘Older’ Folk Pace Resurging Ticket Sales.” Variety, July 9. Golden, Herb. “Hollywood Yens Some Lovin’: Not Easy to Get Boy-Girl Stories.” 1951. Variety, June 13. Goldman, William F. 1983. Adventures in the Screen Trade: A Personal View of Hollywood and Screenwriting. New York: Warner Books. “Gotta Be Smash Hits, Or Else: No In-Betweens on Pic Product.” 1951. Variety, May 30. Gouldner, Alvin. 1954. Patterns of Industrial Bureaucracy. New York: Free Press. Grazian, David. 2004. “The Production of Popular Music as a Confidence Game: The Case of the Chicago Blues.” Qualitative Sociology 27: 137-158. Grindstaff, Laura. 2002. The Money Shot: Trash, Class, and the Making of TV Talk Shows. Chicago and London: University of Chicago Press. Guetzkow, Joshua, Michèle Lamont, and Grégoire Mallard. 2004. “What is Originality in the Social Sciences and the Humanities?” American Sociological Review 69:190-212. Hallett, Timothy and Marc Ventresca. 2006. “Inhabited Institutions: Social Interactions and Organizational Forms in Gouldner’s Patterns of Industrial Bureaucracy.” Theory and Society 35: 213-236. 239 Handel, Leo. 1950. Hollywood Looks at Its Audience: A Report of Film Audience Research. Urbana-Champaign: University of Illinois Press. Haveman, Heather A. and Hayagreeva Rao. 1997. “Structuring a Theory of Moral Sentiments: Institutional and Organizational Coevolution in the Early Thrift Industry.” American Journal of Sociology 102: 1606-1651. Hebdige, Dick. 1979. Subculture: The Meaning of Style. New York and London: Routledge. Hilgartner, Stephen. 2000. Science on Stage: Expert Advice as Public Drama. Stanford: Stanford University Press. Hochschild, Arlie Russell. 1983. The Managed Heart: The Commercialization of Human Feeling. Berkeley and Los Angeles: University of California Press. -----. 1979. “Emotion Work, Feeling Rules and Social Structure. American Journal of Sociology, 85, 551-575. Igo, Sarah E. 2007. The Averaged American: Surveys, Citizens, and the Making of a Mass Public. Cambridge and London: Harvard University Press. Kaplan, Mike. 1951. “Exhibs Sour on Formula Pix: Want ‘Family’ Entertainment.” Variety, June 21. -----. 1950. “Stix Exhibs Cry for New Stars: Gotta Cater to New Generation.” Variety, June 14. Klinenberg, Eric. 2002. Heat Wave: A Social Autopsy of Disaster. Chicago: University of Chicago Press. Johnson, T. 2006. “Any Yen for Ovitz Den?” Variety, 21–27 August, pp. 1, 41. Klady, Leonard. 1993. “Tracking a Lost Generation: Stunted Growth of Core Audience Frustrates Filmers.” Variety, December 5. Klain, Stephen. 1983. “Prods. Over-Value U.S. Art Mart: ‘Classics’ Eye Upfront Stakes as Terms Stiffen.” Variety, May 4. Knorr-Cetina, Karin. 1999. Epistemic Cultures: How the Sciences Make Knowledge. Cambridge, MA: Harvard University Press. Kuhn, Thomas S. 1962. The Structure of Scientific Revolutions. Chicago and London: University of Chicago Press. 240 Lampel, J., T. Lant, and J. Shamsie. 2000. “Balancing Act: Learning from Organizing Practices in Cultural Industries.” Organization Science 11(3): 263-269. Landry, Robert J. 1970. “Theatres Fear of Violence: ‘Revolution’ on Worry Agenda.” Variety, May 13. Latour, Bruno and Steve Woolgar. 1979. Laboratory Life: The Social Construction of Scientific Facts. Beverly Hills and London: Sage. Levy, Emmanuel. 1999. Cinema of Outsiders: The Rise of American Independent Film. New York: New York University Press. Lichterman, Paul. 2005. Elusive Togetherness: Church Groups Trying to Bridge America’s Divisions. Princeton: Princeton University Press Magnuson, Eric. 2005. “Cultural Discourse in Action: Interactional Dynamics and Symbolic Meaning.” Qualitative Sociology, 28, 371-398. Marich, Robert. 2005. Marketing to Moviegoers: A Handbook of Strategies Used by Major Studios and Independents. New York and London: Elsevier. Mauss, Marcel. 1967. The Gift: Forms and Functions of Exchange in Archaic Societies. New York: Norton. Mayer, Vicki. 2008a. “Guys Gone Wild? Soft-Core Video Professionalism and New Realities in Television Production.” Cinema Journal 47(2): 97-116. -----. 2008b. “Studying Up and F**king Up: Ethnographic Interviewing in Production Studies.” Cinema Journal 47(2): 141-148. Meyer, John W. and Brian Rowan. 1977. “Institutionalized Organizations: Formal Structure as Myth and Ceremony.” American Journal of Sociology 83: 340-63. Moody, Michael and Laurent Thévenot. 2000. “Comparing Models of Strategy, Interests, and the Public Good in French and American Environmental Disputes.” In Michèle Lamont and Laurent Thévenot, Rethinking Comparative Cultural Sociology: Repertoires of Evaluation in France and the United States. Cambridge: Cambridge University Press. Morgan, David L. 1988. Focus Groups as Qualitative Research. Newbury Park: Sage. MPAA. 2007. “2006 U.S. Movie Attendance Study.” Motion Picture Association of America. Retrieved online at www.mpaa.org. 241 -----. 2006. “2005 U.S. Movie Attendance Study.” Motion Picture Association of America. Retrieved online at www.mpaa.org. Napoli, Philip M. 2003. Audience Economics: Media Institutions and the Audience Marketplace. New York: Columbia University Press. “NATO-MPAA Eye ‘Missing’ Public.” 1970. Variety, March 11. Ohmer, Susan. 2006. George Gallup in Hollywood. New York: Columbia University Press. Peterson, Richard A. and N. Anand. 2004. “The Production of Culture Perspective.” Annual Review of Sociology 30: 311-34. Pettera, A., Ed. 2006. Zagat Survey: Los Angeles/Southern California Restaurants, 2007. Zagat Survey. Pierce, Jennifer L. 1995. Gender Trials: Emotional Lives in Contemporary Law Firms. Berkeley and Los Angeles: University of California Press. Powdermaker, Hortense. 1950. Hollywood: The Dream Factory: An Anthropologist Looks at the Movie-Makers. Boston: Little, Brown and Company. Pryor, Thomas M. 1970. “Crucial Summer for Film iz: H’wood Future Riding on B.O.” Variety, July 1. Ross, Steven J. 1998. Working Class Hollywood: Silent Film and the Shaping of Class in America. Princeton, New Jersey: Princeton University Press. Rosten, Leo C. 1941. Hollywood: The Movie Colony, the Movie Makers. New York: Harcourt, Brace and Company. Schacht, S. P. 1996. “Misogyny On and Off the ‘Pitch’: The Gendered World of Male Rugby Players. Gender & Society, 10, 550-565. Scott, Allen J. 2005. On Hollywood: The Place, the Industry. Princeton and Oxford: Princeton University Press. -----. 2004. “The Other Hollywood: The Organizational and Geographic Bases of Television-Program Production.” Media, Culture & Society 26:183-205. Scott, Tony. “So Who’s Going For Smut? When Fans Ask, ‘Why the Dirt?’” 1970. Variety, August 12. Setlowe, Rick. 1970. “Grim Era Needs Grin Pix: Folks in Mood For Love, Gags.” Variety, August 19. 242 Sewell, William F. 1992. “A Theory of Structure: Duality, Agency, and Transformation.” American Journal of Sociology 98: 1-29. Silverman, Syd. 1970. “Key to Film Box Office Graphs.” Variety, January 14. Simmel, Georg. 1950. The Sociology of Georg Simmel. Trans. K. H. Wolff. New York: The Free Press. Sklair, Leslie. 2001. The Transnational Capitalist Class. Malden, MA: Blackwell. Smythe, Dallas. 1981. "On the Audience Commodity and Its Work." In Dependency Road: Communications, Capitalism, Consciousness, and Canada, Pp.22-51. Norwood, NJ: Ablex. Snyder, Gabriel. 2006. “Sorry, Wrong Number: Screwy Tracking Scores Leave Studios Vexed.” Variety. July 24-30: 1, 49. Spilker, Eric. 1970. “Majors Staff Up for Youth: Assign Contacts to Cover Field.” Variety, June 24. Srinivas, Lakshmi. 2003. "The Active Audience: Spectatorship, Social Relations and the Experience of Cinema in India." Media, Culture & Society 24(2): 155-73. Sternheimer, Karen. 2006. Kids These Days: Facts and Fictions About Today’s Youth. New York: Rowman & Littlefield. Swidler, Ann. 2001. Talk of Love: How Culture Matters. Chicago and London: University of Chicago Press. Taylor, Thom. 1999. The Big Deal: Hollywood's Million-Dollar Spec Script Market. New York: William Morrow and Company, Inc. Thévenot, Laurent. 2001. “Organized Complexity: Conventions of Coordination and the Composition of Economic Arrangements.” European Journal of Social Theory 4: 405-425. “Top Pix for Summer Release: 95 Films Due to Fight Hot Mos.” Variety, May 17. “Try Anything as B.O. Panacea: Exhibs Getting More Desperate.” 1950. Variety, May 24. Tuchman, Gaye. 1978. Making News: A Study in the Social Construction of Reality. New York and London: The Free Press. 243 Turow, Joseph B. 1997. Breaking Up America: Advertisers and the New Media World. Chicago: Chicago University Press. U.S. National Center for Health Statistics. 2001. “Table 1. Live Births, Birth Rates, and Fertility Rates, by Race: United States, 1909-2000.” Retrieved online at http://www.cdc.gov/nchs/data/statab/t001x01.pdf. “U.S. Youth Threat.” 1970. Variety, May 6. Van Maanen, John. 1988. Tales of the Field: On Writing Ethnography. Chicago and London: University of Chicago Press. Verrill, Addison. 1970a. “Oracles Edgy, Buffs Scoff: As Old Format Films Big B.O.” Variety, April 15. Verrill, Addison. 1970b. “U.S. ‘Doomsday’ as Pix Kick: Gotta Be Bleak in Underground.” Variety, June 17. Vogler, Christopher. 1993. The Writer’s Journey: Mythic Structure for Writers. New York: Wiese, Michael Productions. Wasko, Janet. 2003. How Hollywood Works. Thousand Oaks and London: Sage. West, Candace and D.H. Zimmerman. 1987. “Doing Gender.” Gender & Society 1: 125- 151. Williams, Christine. 2006. Inside Toyland: Working, Shopping, and Social Inequality. Berkeley and Los Angeles: University of California Press. -----. 1989. Gender Differences at Work: Women and Men in Nontraditional Occupations. Berkeley and Los Angeles: University of California Press. Wyatt, Justin. 1994. High Concept: Movies and Marketing in Hollywood. Austin: University of Texas Press. Young, Jock. 1970. The Drugtakers: the Social Meaning of Drug Use, London: McGibbon and Kee. Zafirau, Stephen. 2008. “Reputation Work in Selling Film and Television: Life in the Hollywood Talent Industry.” Qualitative Sociology 31: 99-127. -----. 2009. “Audience Knowledge and the Everyday Lives of Cultural Producers in Hollywood.” In Local Production Studies. Edited by Vicki Mayer, John Caldwell, and Miranda Banks. New York: Routledge. 244 Appendix A: Studying Hollywood Moviemaking As an ethnographer doing research in Hollywood, I have sometimes thought of a favorite cartoon (published years ago) that pokes fun of the contemporary ethnographic enterprise. The cartoon depicts an anthropologist visiting a hunter-gather tribe whom he is studying, and the caption quotes the anthropologist as saying to the tribespeople: “But enough about me, what about you?” It has now become conventional for ethnographers to write “confessionals” on their experiences in the field (Van Maanen 1988). As the above cartoon points out, this act of confessing can potentially be overdone and perhaps even border on self-indulgence. But putting these risks aside, such confessionals can also have important value for an ethnographic study. They can shed light on how an ethnographer’s presence in the field helped construct the social setting in which ethnographic data was gathered, and, ultimately, the backstory of how the author arrived upon conclusions. Given my critical focus on the social conditions under which knowledge is produced in Hollywood, it is only fitting that this same critical eye be applied to (albeit in a self-reflexive way) to the knowledge I have attempted to advance here. This is necessary in order to bring attention not only to my study’s inevitable limitations but also any methodological strengths. Additionally, this exercise is also potentially valuable to other ethnographers. I see doing ethnography as a kind of systematic art, and ethnographers collectively benefit from hearing other accounts of the art of ethnographic research. The specifics of my experiences as a researcher may add to the accounts of others doing ethnographic work on the “culture industries” as well as those doing work in other areas (even topically different ones). 245 Participant observation research eventually results in a written account of social life, and such accounts are always built upon a series of choices that the researcher necessarily makes as she works in the “field.” These choices involve a variety of questions, including: Which sites should be studied? How does one present oneself to those being researched? What sorts of “bargains” and “compromises” does one make as she gains access to exclusive settings? What, among the apparent chaos of social life, does the participant observer look for once she is in those settings? And, theoretically speaking, how does the ethnographer know what he or she has a “case” of? Participant observation researchers have a number of different strategies for dealing with these issues. For me, the answers to these questions were guided by one particular approach to ethnographic research, what Michael Burawoy calls the “extended case method” (Burawoy 1999). The extended case method is essentially a theory-driven strategy for carrying out ethnographic research, one in which a participant observer begins with preexisting theoretical concepts and seeks to advance/reconstruct them through systematic participant observation. This approach contrasts, for example, with the ethnogrpahic method advocated by “grounded theory” (Glaser & Strauss 1967), which proposes that researchers should enter the field free of preconceptions and gradually builds concepts (and eventually an ethnographic account) from these “grounded” observations. Given problems involved in attempting to enter the field without preconceptions (namely, the extent to which the researcher’s own subjectivity and positionality might shape observations), the extended case method offers an alternative approach to ethnography in which weaknesses often associated with ethnographic research are harnessed as strengths. Here, preexisting theory becomes the guide for the 246 entire ethnographic research process, from site selection, to forming ethnographic hypotheses, to eventually writing an account of the social world. Research proceeds by first attempting refute existing theoretical explanations (in whole or in part) and then reconstructing them as stronger explanations. The body of theories utilized in this process can be quite broad, ranging from “a stock of academic theory on the one side and existent folk theory or indigenous narratives on the other” (Burawoy 1999: 7). In this way, the extended case method is a theory and method of ethnographic research that is fundamentally different from (though complimentary to) “postitivist” science. Unlike science based on the postivist model, ethnographic research that utilizes the extended case method is based on a “reflexive” model of science that need not be concerned with issues of representativess, relaibility, or replicabilty. Only a single case is needed to engage with and reconstruct preexisting theory. The Extended Case Method Extended to Hollywood For me, I began this project with a broad question: How do moviemakers understand their audiences? In other words, dow do they determine audience tastes from the socially isolated enclave of the Hollywood entertainment industry? And, how do these conceptions of the audience become meaningful in the process of making popular film content? In short, I was interested in how producers related to their audiences. The “stock” of theory I began with involved a number of sociological theories of “cultural production,” including explanations offered by neoclassical economics, various interpretations of the mass media emanating from Marxism/critical theory, Bourdieu’s understanding of cultural production as organized within a hierarchical “cultural field,” 247 the “production of culture” school, as well as the entertainment industry’s own “folk theories” of itself . These perspectives all offered different explanations for how “the audience” factored into production, ranging from seeing producers as having conscious intentions to manipulate audiences to seeing producers as almost completely hapless when it came to determining audience tastes. I began fieldwork at Limelight Entertainment in the fall of 2005 attempting to gather grounded observations that would speak to (and possibly refute) these preexisting theoretical perspectives. Knowing which theories to engage with became clear based on my observations. Applying some of these theoretical insights (e.g., neoclassical economics, some interpretations of Marxism) to my particular observations at Limelight would have involved heavily superimposing these perspectives on my observations. Beginning with my most grounded observations at Limelight, however, I began to systematically code several recurring realities. First, life in this Hollywood talent management/production firm was a special kind of “scene” (in the Goffmanian sense), with special rules for interaction and self-presentation. My attendance at a business school course on the “talent representation” business supported the idea that this type of “scene” was not particular to Limelight, but the talent industry more broadly. Second, there was a particular language used at Limelight for evaluating movie ideas (what I eventually came to call an “intuitive discourse” of audience appeal). Again, this way of speaking about movie ideas did not appear to be confined to Limelight, but seemed to be the way in which agents/talent managers I encountered more broadly tended to speak about audiences. Also noteworthy was the fact that there was another sort of discourse that I 248 heard used elsewhere in the entertainment industry, but that was habitually avoided at Limelight. This was what I eventually called “technical discourses” of audience appeal. Thus, I noticed both patterned ways of interacting and patterned ways of “sensemaking” about audiences within the Hollywood talent industry. Given that these patterns seemed to operate as rules that were transposable across specific organizational sites, I was led to neoinstitutional organizational analysis as a theoretical starting point. And, there were some studies of the “culture industries” that had preceeded me in using the notion of “institutions” to analyze media production (Ahlkvist & Faulkner 2002; Bielby & Bielby 1994). A common theme in these handful of studies was that institutions proliferate in culture industries in response to deep, underlying uncertainties about what sorts of content will “work.” Studies such as these picked up on neoinstitutionalism’s general concern over the importance of uncertainty/ anxiety in producing organizational behavior. However, my observations at Limelight and elsewhere had provided me with some theoretical puzzles. First, the lived uncertainties of work at Limelight were as much about the relationships with other producers as they were about uncertainties about the product. Evaluations of movie ideas were spoken about in confident, even over-confident terms. Relationships and industry status, however, were experienced as tenuous, as evidenced by the great deal of energy that managers at Limelight devoted to maintaining their reputations for others in the industry. It was here, in their relationships with other “players” in the industry, that talent managers’ lived anxieties seemed most apparent. A second puzzle involved how “technical discourses” about audiences were (not) used at Limelight Entertainment. These “technical discourses” about audiences appeared to provide immense certainty 249 about movie ideas, being that this language was couched in notions of “objectivity.” Why, I wondered, in a world in which “nobody knows anything” about which movie ideas would work, did Limelight’s managers systematically push aside this technical language and opt for seemingly more intuitive ways of evaluating movie ideas? These sorts of puzzles are key moments in the ethnographic process because they lead to new insights. In my case, I began to form a larger hypothesis about how Limelight worked as a site in which knowledge about audiences was reproduced. Specifically, I hypothesized that the institutionalized way in which Limelight evaluated movie ideas and conceptualized audiences was embedded in the institutionalized set of rules that governed interaction within the talent industry more broadly. That is, I guessed that this particular discourse about audience appeal (the intuitive one) became useful within the strategic interactional games that Limelight’s managers had to play within their industry, while the technical disocourse of audience appeal became problematic in this world. Perhaps, I wondered, intuitive discourses were being used as a way of navigating the potential perils of work in the talent industry. Although this hypothesis appeared to make sense, it could ultimately only be validated through observing a comparison case. This led me to look for another site within Hollywood in which audience knowledge would be reproduced. Ideally, this comparison site would vary one of the two larger “variables” in my working hypothesis: “type of discourse about audience” or “interactional context.” Since I already knew that there was a competing “technical discourse” of audience appeal that was used by moviemakers in some quarters, I became interested in finding a site where this sort of knowledge was reproduced. This pushed me to seek access at an audience research firm 250 or research wing of a motion picture studio, and eventually led to me using Strategic MediaVision as a comparison site. At MediaVision, I immediately began looking for aspects of work that could be coded as “interactional context” and “discourse about audiences.” Here, I saw a different kind of audience discourse (a “technical” one, as expected), but was surprised to see exactly how this type of audience discourse functioned in the world of Strategic MediaVision. In contrast to Limelight, work at MediaVision was much more routinized and the interactional context (and its attendant anxieties) was heavily defined by this profit-driven routinization. Here, predictable, convenient stories about audiences (told in the language of social scientific research) were used as researchers attempted to maintain the appearance of objectivity. At both Limelight Entertainment and Strategic MediaVision, interactional rules that governed everyday work and the kind of knowledge about audiences that was being reproduced worked in tandem. Workers at both sites strove to affirm reputations based on different premises: at Limelight, reputation was premised on being perceived as a power player in Hollywood, and at MediaVision it was premised on appearing “objective” and technically competent. The type of knowledge being reproduced at each site thus serviced different sorts of interactional games. In this way, this comparison provided a way to extend previous organizational analyses of cultural production. Access Hollywood: Methodological Challenges At the beginning of this Appendix, I alluded to a cartoon depicting an anthropologist who, after what we are to infer is a longwinded introduction of himself and his purpose to the tribe he is studying, finally gets around to asking the tribespeople about themselves. 251 This cartoon also became relevant for me as I carried out this project because it came to represent the inverse of my own experience as an ethnographer. I ended up spending very little time talking with “the researched” about my role as a researcher. This was not because I was doing seriptitious research, as all of those I interacted with at Limelight and MediaVision ultimately knew I was carrying out a research project at a university. It had more to do with the many assymetries that existed in the sites I studied, and my status at the bottom of these hierarchies. As the cartoon mentioned above suggests, contemporary ethnographers are tremendously reflexive about their own power over those they write about. But, in my case, I found this power dynamic largely reversed. For the vast majority of my time at each site, those I observed seemed profoundly disconcerned with the fact that I was doing a research project and much more concerned with the everyday rhythms of work and my ability to function competently in them. My experience felt similar to the one decribed by Laura Grindstaff during her particpant observation study of television talk shows: access proved to be an ongoing, incomplete process as I attempted to study those “higher” up in the entertainment industry. And, for much of my time in the field, I became largely invisible to those working at more senior levels (although this was much more the case at Limelight than MediaVision) (see Grindstaff 2002). Ethnographers of the culture industries frequently cite significant challenges gaining entry into the worlds of media production (if an ethnographic approach is indeed successful at all) (see, for instance, Gridstaff 2002; Gamson 1998). It is easy to speculate why this is the case: moviemakers, for example, experience little foreseeable benefit to allowing researchers in, and they can perceive many possible risks (real or not)—not the 252 least of which is the time involved in participating in academic research. Like all ethnographers, then, researchers who study media production must be strategic in how they approach potential gatekeepers and how present themselves to informants. And, in order to gain access, they must often be willing to strike an “ethnographic bargain” with those being studied. Ethnographic research on the culture industries can thus prove to be a very imperfect process, demanding flexibilty and compromise on the part of the researcher. Vicki Mayer put this well when she reflected on her study of soft-core video production: Studying up, as described in this case, involved f**king up, not only in terms of my errors in judgment, but also in terms of subtle machinations of power and pleasure. Like the tired anecdote of the aspiring actress who sleeps her way to the top, I felt both a sense of being a user and being used. (2008b: 145) In this inevitable, and at times uncomfortable game of “using and being used,” researchers must trade on the resources they have at their disposal. My lack of key resources immediately presented some problems when I began this project. Namely, I lacked any sort of usable connection to those working in the Hollywood entertainment industry (except for a neighbor who was forever trying to sell his television pilot). Matters of appearance also weighed on my mind as I conceptualized an ethnographic or even an interview study. Would it really work if I were to roll up to a sleek talent agency or a studio lot in Beverly Hills, complete with valet parking, driving my 1987 Nissan Sentra? And, how many Hollywood-worthy outfits did I currently own? In the end, by borrowing my fiance’s newer car and making several trips to the mall, I was ultimately able to adjust to the respective aesthetics of work at Limelight and MediaVision. 253 However, gaining access to the places and situations I wished to observe remained an issue. To overcome this, I was fortunate to be able to draw on resources that I did have. For one, I was a graduate student at the University of Southern California, which has very close connections (along with UCLA) to the Hollywood film industry. I attempted to take advantage of some these connections by taking part in programs offered through USC’s School of Cinematic Arts, and eventually was able to position myself to have interviews for internships. Despite the fact the I was working on a Ph.D. in Sociology, my affiliation with USC provided a kind of “cultural capital” that I could draw upon when attempting to gain these internships. A second resource I was able to draw upon was my available time and my willingness to do the work of being an intern at a Hollywood talent firm. At Limelight, this was part of the “ethnographic bargain” I struck: I performed without pay a large quantity of the very mundane tasks it took to help Limelight function (photocopying, pagechecking scripts, faxing, running packages and other odd errands, stocking beverages), and I was granted admission to a Hollywood talent firm that represented A- list stars and worked major deals in the film business. The fact that I was in my mid- twenties at the time and a student probably helped a great deal in assuming this identity. With a Master’s degree, I was told that I was “overqualified” to work at Limelight as an intern, but it nevertheless seemed natural for me to be at Limelight once I was in the door. I would appear to be another twentysomething trying to establish himself in the entertainment industry. 254 Interning at Limelight Entertainment At Limelight, I, like all interns, strove to learn the business from the ground up. As an ethnographer, this at times became frustrating because it meant putting in a great deal of time before I felt I was gaining meaningful insights. On many days, the observations I made felt “thin.” But I speculate that my hard work and attempts to do my job efficiently at Limelight allowed people at the company to feel indebted to me and facilitated me building usable relationships. I ultimately worked at Limelight several full days a week over a period of seven months. During that time, I did a great deal of the day-to-day tasks needed to operate a talent firm, including photocopying, running errands, keeping supplies stocked. Eventually, I was allowed to perform more extensive duties, substituting for managers’ assistants phone duties and reading and covering scripts. Gradually, I became less invisible at the firm while Limelight’s workers became more ingratiated toward me. This first became evident when the Holiday giftgiving season came around, and I found that the partners of the company had all pitched in to get me an iPod Nano as thanks for my hard work over the previous months. The assistant who gave it to me explained, “You work so hard… well, for nothing! So, the partners wanted to pitch in and get you this.” A couple of months after that, I began to get offers to advance up the ranks in the company. A job ad for an upcoming vacancy as one of the managers’ assistants was sent to my email account. Shortly after that, I was offered a job as a production assistant on a film project that Limelight was setting in motion, and would have resulted in me working full-time in Vancouver. Taking either of these 255 positions would not have fit with my goals as a researcher, nor would it have been practically feasible. If these offers revealed the extent to which I was being trusted in the company and viewed as compentent, they also revealed that my aims as a researcher were incongruent with how those in the company likely viewed me. To them, I was primarily a slightly more mature USC student who was trying to break into the entertainment business. At times, this inconcruency presented me with some ethical discomfort, and at various points throughout my time at Limelight, I tried to make my presence as a researcher known (without trying to come on too strong). During my initial interview at Limelight, I told the people interviewing me that I was “writing my thesis” on the Hollywood talent industry and that I would like to “observe” Limelight as part of my study. A couple of months into my internship, I sent a detailed email to the assistants of Limelight’s partners describing my project and the nature of ethnography, how I was collecting observations, and how I would ultimately use the observations (in the form of publications, conference presentations, etc.). Finally, in the process of gaining formal permission to be at Limelight, I eventually sent an email to one of company’s partners describing my project in detail. I did all of this in the interest of clearing up any misconceptions about what my goals at the company were, but also in the hopes that my level of access might be heightened (a goal that ran up against limits). I ultimately worked at Limelight two or three full work days a week over a period of seven months. During that time, I gradually gained access to more of the action at Limelight: I was able to observe more and more situations, developed relationships with many at the company, and offered access to a variety of documents (screenplays, 256 coverage memos, etc.). However, my access to those spaces where the company’s managers spoke informally with their clients, other managers, or business associates was much more limited. At one point (a few months into fieldwork), I asked one of Limelight’s partners if I could sit in on more senior staff meetings where I knew ideas about movie projects were discussed, a request that was refused. My goals as a researcher, then, ultimately ran up against limitations supplied by the norms and priorities that managers at the company had for conducting business. The bargain struck between Limelight as research site and me as a researcher stemmed from two competing sets of priorities, and ultimately resulted in useful, but only partial access to the world I sought to study. Interning at Strategic MediaVision I entered Strategic MediaVision by applying for a position as an intern, and brought with me the credentials of being an advanced Ph.D. student at the University of Southern California who was familiar with social scientific research, statistics, and also writing a dissertation on “the significance of audience research in the Hollywood film industry.” When I interviewed for this position, I told a company’s human resources officer that I would like to use my experience working at MediaVision as part of my research and provided her with a concise written statement of my aims as a researcher. I also described my project to Martha when I interviewed with her, and, shortly after I was hired, sent around an email to the various members of my workgroup telling them that I was working on a doctoral dissertation and intended, with their permission, to observe how research was conducted at the company. The world of academia and writing theses 257 was a familiar those I directly worked with and observed (some of whom came from academia themselves), and getting their permission to observe the research process at the company turned out to be easier than I anticipated. The data that I ultimately gathered included a month of observations, three full days a week. While this is perhaps a short time span for an ethnography, the data I was able to gather was very rich—especially when compared to Limelight. While my experience at Limelight was one in which I was initially very invisible, at MediaVision I was immediately integrated into the the company’s research operations. I was immediately included in workgroup meetings and participating in the production of audience reports. Thus, although the quantity of time I spent at MediaVision was much shorter (amounting to approximately 100 hours of observation), I was still able to rapidly find patterns in the research process at the company.
Abstract (if available)
Abstract
Many scholars of the mass media highlight how the production of popular culture in contemporary media industries powerfully involves notions of "the audience." That is, media producers draw upon collective assumptions surrounding who audiences are and what sorts of content they prefer as they go about making media content. This study proposes and develops an "evaluation in interaction" approach to examine how moviemakers socially construct American movie audiences within the context of their everyday work in the Hollywood film industry. Specifically, I show how moviemakers construct audiences using two institutionalized discourses of evaluation -- both "intuitive" and "technical" -- that are prevalent in the Hollywood film industry. This dissertation uses archival data to document the emergence and institutionalization of both intuitive and technical discourses of audience appeal over the latter half of the twentieth century. Then, drawing on organizational ethnographies of both a Hollywood talent firm and an audience research firm, as well as interview data with film industry professionals, I show how intuitive and technical discourses of audience appeal are reproduced within everyday work settings of the Hollywood film industry. Centrally, I argue that both intuitive and technical discourses about American motion picture audiences are embedded within the particular organizational conditions in which they are reproduced.
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Tacit anticipation among film students: an ethnography of making movies in film school
PDF
Advocacy of public interest content in corporate entertainment media
PDF
Effects of economic incentives on creative project-based networks: communication, collaboration and change in the American film industry, 1998-2010
PDF
The Hollywood left: cinematic art and activism in the 1930s
PDF
Celestial bodies: Black women, Hollywood, and the fallacy of stardom
PDF
Bad boys, reform school girls, and teenage werewolves: the juvenile delinquency film in postwar America
PDF
The data dilemma: sensemaking and cultures of research in the media industries
PDF
The informationalization of race: communication technologies and genomics in the information age
PDF
Comic books incorporated: industrial strategy and the legitimation of lowbrow media
PDF
Image breakers, image makers: producing race, America, and television
PDF
The joy of censorship: strategies of circumvention in novel and film
PDF
Hollywood's 2015 diversity problem and the powerful impact of film and television
PDF
One of us: exploring the relationship between Hollywood and disability
PDF
These are their stories: two decades of Showrunner production, content, and context in Law & Order: SVU
PDF
Aftermarkets of empire: South Korean popular music and global logics of race and gender in the U.S. media industries
PDF
Toward counteralgorithms: the contestation of interpretability in machine learning
Asset Metadata
Creator
Zafirau, Stephen
(author)
Core Title
Imagined audiences: intuitive and technical knowledge in Hollywood
School
College of Letters, Arts and Sciences
Degree
Doctor of Philosophy
Degree Program
Sociology
Publication Date
01/24/2009
Defense Date
08/06/2008
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
cultural production,evaluation in organizations,Hollywood,media audiences,OAI-PMH Harvest,social construction of knowledge
Place Name
Hollywood
(city or populated place),
USA
(countries)
Language
English
Contributor
Electronically uploaded by the author
(provenance)
Advisor
Eliasoph, Nina (
committee chair
), Clarke, Peter (
committee member
), Lichterman, Paul (
committee member
)
Creator Email
szafirau@tulane.edu,zafirau@usc.edu
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-m1955
Unique identifier
UC1192613
Identifier
etd-Zafirau-2570 (filename),usctheses-m40 (legacy collection record id),usctheses-c127-153685 (legacy record id),usctheses-m1955 (legacy record id)
Legacy Identifier
etd-Zafirau-2570.pdf
Dmrecord
153685
Document Type
Dissertation
Rights
Zafirau, Stephen
Type
texts
Source
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Repository Name
Libraries, University of Southern California
Repository Location
Los Angeles, California
Repository Email
cisadmin@lib.usc.edu
Tags
cultural production
evaluation in organizations
media audiences
social construction of knowledge