Constructing Credible Images: Documentary Studies, Social Research, and Visual Studies Wagner. Jon The American Behavioral Scientist; Aug 2004; 47, 12; ProQuest Central pg. 1477 Constructing Credible Images Documentary Studies, Social Research, and Visual Studies JON WAGNER University of California, Davis Documentary photographers and visual social researchers both rely on visual, real-world data but differ in how they address three key challenges: creating empirically credible images of culture and social life, framing empirical observations to highlight new knowledge, and challenging existing social theory, A close look at these shared challenges and contrasting conventions recommends exemplary documentary studies as an empirically viable alternative to traditional social research. Three recent projects illustrate the value of such alternatives to informing public discourse. They also embody strategies that could enhance the value of empirical, visual inanity to education and other community-based fields of practice. Keywords: documentary: visual; research; photography; sociology Scholars interested in culture and social life have puzzled for some time over how to draw the line between social research and documentary studies, documentary photography, and filmmaking in particular. Some social scientists clearly embrace documentary studies as a vital complement to their own work, but others dismiss them for a lack of rigor and depth or for neglecting social theory in favor of anecdotes, evocations, and pretty pictures. Similarly, although some image makers regard the social sciences as a valued foundation for documentary photography and filmmaking, others find them overly abstract and impersonal, insensitive to a fault, pedantic, or beside the point. For the past 30 years or so, the terms defining this divide have been shaped for many social scientists by two seminal and complementary statements: John Collier Jr.'s (1967) Visual Anthropology: Photography as a Research Method provides a thoughtful and encouraging account of how photographs could be used to make durable, visual records of culture and social life and to interview research participants through a process of "photo elicitation." Though he had worked previously as a documentary photographer in the Farm Security Administration (see Figure 1), Collier's monograph argues that the kind of image making most appropriate to the social sciences was systematic, deliberate, and well articulated with a traditional research design. Howard Becker's (1986b) essay AMERICAN BEHAVIORAL SCIENTIST, Vol. 47 No. 12, August 2004 1477-1506 DOI: 10.1 177/0002764204266235 O 2004 Sage Publications 1477 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1478 AMERICAN BEHAVIORAL SCIENTIST Figure 1: A photograph made by John Collier Jr. when he worked for the Farm Security Administration. The original caption reads "Red Cross distributing knitting material. San Francisco, California, 1941." Photo courtesy of the Library of Congress. "Photography and Sociology" (originally published in 1974) explicitly sets aside the kind of systematic recording Col Her recommended and makes a somewhat different argument: that social documentary photography itself shared important elements of inquiry and representation with sociological work. In this article I will propose a framework that takes the two kinds of image making that Becker (1986b) and Collier (1967) teased apart and brings them together within a shared rubric of empirical social inquiry. Within this framework, documentary photography and visual social research are distinguished not so much by different logics of inquiry as by conlrasting social conventions for addressing three key challenges: creating empirically credible images of culture and social life, framing empirical observations to highlight new knowledge, and challenging existing social theory. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Wagner/CONSTRUCTING CRI-DIBI.E IMAGES 1479 In presenting (his case, 1 will first describe what I mean by empirical inquiry, visual social inquiry in particular, and where that fits within documentary studies and social scientific practice. I will then review three recent documentary projects, paying particular attention to how they address the inquiry challenges noted above. A close look at these projects reveals several taken-for-granlcd features of social research that warrant further inquiry. These include the preeminence of research designs over personal accounts of observation and recording, a reliance on academic communities in defining new knowledge, and an overweening attraction to explicit, rather than implicit, statements of social theory. Treating these features as working conventions of professional social researchers, rather than as determinants of systematic, empirical inquiry per se, could encourage more back and forth between social research and documentary studies. Taking that prospect seriously could stimulate new kinds of visual studies and enhance the value of empirical, visual inquiry in education, community development, and public discourse. CHARTING THE EMPIRICAL DIVIDE For the purposes of this article, I will define empirical social inquiry as an effort to generate new knowledge of culture and social life through the systematic collection and analysis of sensory evidence and other forms of real-world data. This definition falls across and somewhat outside the conventions of both academic social science and professional documentary work. It contrasts, for example, with the narrow view held by some social researchers that empirical studies are necessarily quantitative. It also contrasts with the convictions of some image makers that social advocacy, artistic vision, and technical skill are, by themselves, enough to construct empirically sound images. Both social researchers and lay readers typically find images of culture and social life to be more credible when they are based on extensive and detailed observation in an appropriate array of natural settings, backed up by other data, and presented in ways that invite analysis, including commentary from the people they depict. But photographs can support empirical inquiry in ways that do not always square with popular notions of what makes them "true" or "false,'' and there is a danger in trying to turn these notions into categorical prohibitions or ideals. For example, we are understandably suspicious of photographs that reflect contrived poses or processing distortions or that come with captions that misrepresent an image's origins or typicality. But posed photographs provide valuable evidence of how people want to be seen by others (Pinney, 1997; Ruby, 1995), and photographed reenactments (see Figure 2) can generate credible visual records not otherwise available (Kroeber, 2002; Ricger, 2003). Similarly, although page layouts featuring severely cropped and juxtaposed images can Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 2: Using a "peeling spud" made from an old car spring, semiretired logger Ernie Toivonen demonstrates, in 1990, the hand craft of debarking a tree in Ontonagon County, Michigan. In the course of an interview with the sociologist Jon Rieger, Toivonen offered to "show him how it was done" prior to mechanization of the pulp-wood industry in the 1970s and 1980s. Following up on Toivonen's invitation to reenact this technique, Rieger photographed aspects of a logger's craft that were no longer practiced and for which no historical images were to be found. Both photographs by Jon Rieger. Wagner / CONSTRUCTING CREDIBLE IMAGES 1481 create false impressions, they can also highlight theoretically significant details and comparisons. These ambiguities are complicated by the routine fabrications of social life, including the social life it takes to conduct empirical inquiry. For example, although evidence can be collected systematically within specific data categories—time-lapse photographs of people crossing a street; census reports of race, income, and ethnicity; or sociometric charts of child friendships—the categories and data collection tools that guide work of this sort are themselves socially constructed. Some people have taken this to mean that we should forget about empirical inquiry altogether, but it leads me to recommend doing somewhat more of it rather than less. It seems a bit foolish, for example, to dismiss a number, chunk of text, or photograph just because it has been "socially constructed." It is difficult to imagine anything that is not! A better bet would be to examine the construction process itself for what we can learn about not only the number, text chunk, or image but also the social contexts in which they are shaped and distributed. Which leads me back to Collier (1967) and Becker (1986b), each of whom affirmed, in complementary fashions, the value of photographs as durable and useful records of what was visible in a particular time and place. It is not always easy to make those records in the way we will later find most useful, nor is it a simple matter to understand what images made by others fairly depict or neglect. Indeed, the idea that photographs and other machine-recorded data can be generated without human agency and choice of any sort is both naive and misguided, but so is the idea that a photographer's selectivity in one dimension makes an image wholly suspect in all others (Schwartz, 1999). With this in mind, the credibility and utility of photographs within empirical social inquiry rests not so much on whether they accurately reflect or arbitrarily invent the real world but on how those aspects of the real world they invent or reflect are related to questions we care about (Becker, 1986a)—to understand that, we need something more than the photograph itself. RESEARCH DESIGNS AND PERSONAL ACCOUNTS One tool for helping researchers and their audiences judge what a set of photographs might contribute to a project of empirical inquiry is the research design, an explicit description of how a study is organized and how the right kind of evidence can be brought to bear' in answering predetermined questions. By and large, that is where social scientists place their own trust and hope. Regardless of the kinds of data they choose to examine, a good research design advances the claim that the researcher has conducted (or is about to conduct) an empirically sound investigation. Reproduced with permission of the copyright owner. Further reproduction prohibited without permi: 1482 AMERICAN BEHAVIORAL SCIENTIST Relying on research designs to advance these claims suggests that the main threats to empirical inquiry are those that a research design can guard against. Toward that end, statements by social scientists frequently do a good job of accounting for sample size, for example, and for site selection, the wording of survey questions, or the preparation of appropriate observation schedules and coding strategies. But social science research designs are typically silent about other potential pitfalls. They rarely note the full range of an investigator's interest in a topic or a study site, or preview indeterminate features of the research process, or describe the researcher's honesty, interpersonal skills, or an ability to elicit cooperation and useful information from research participants. Leaving these potentially problematic elements out of a research design affirms a world in which the researcher's role dominates the researcher, in which research designs transcend observation and inquiry crafts. Personal accounts represent another tool for establishing the credibility of empirical social inquiry. They are used rarely by professional social researchers (though efforts lo clarify a researcher's "posilionality" are of a kindred sort) but frequently by documentary image makers. Some such accounts arc infused within the body of a documentary project itself, the way James Agee spoke for himself and Walker Evans in Let Us Now Praise Famous Men (Agee & Evans, 1939/1960). In other instances, they appear as forewords, afterwards, and in interviews that documentary image makers give about their work (Light, 2000; Loengard, 1998; Lyons, 1966; Morris, 1999). In the aggregate, these narrative accounts by documentary image makers affirm a world in which persons shape inquiry and in which the craft of inquiry transcends the research design. When considered together, research designs and personal accounts reveal multiple and complementary dimensions along which empirical inquiries can be more or less well grounded and well executed. These dimensions apply to issues of data collection and analysis and to many other choices investigators make as they go about their work—deciding when data are complete enough to warrant analysis, selecting details to report as illustrations and examples, choosing a starting point for introducing or framing a study, pitching descriptions to a particular level of abstraction or generality, identifying or cultivating audiences for which a study might be of interest, and so on. By paying attention to how these choices affect the truthfulness of their work, social researchers and documentary image makers stand on the same side of the empirical divide. This separates their work from other ways of approaching the world—divine revelation, for example, or fantasy making, psychological projection, speculation, and demagoguery. Individual social researchers and documentary image makers are never completely free from these contrary inclinations. However, within idealized forms of the work in which they are engaged, the latter appear as liabilities, shortcomings, and failures. The logic of empirical inquiry requires that they be addressed, at times by avoiding inventions of the investigator, at other times by bringing them into play. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Wagner / CONSTRUCTING CREDIBLE IMAGES 1483 Figure 3: Photo by Walker Evans, courtesy of the Library of Congress. The original caption reads, "Movie theatre on Saint Charles Street. Liberty Theater, New Orleans, Louisiana. 1935-36." INVKN HONS AND RKKLKCTIONS Social researchers, photographers, and artists can be more or less self-conscious about what their accounts and reports have added to what they have seen. Walker Evans, for example, referred to his work not as documentary photography but as "art done in a documentary style" (Hambourg, Rosenheim, Eklund, & Fineman, 2000). Other documentary image makers have been less careful or held contrary beliefs. In his prejudicial framing, selection, and printing of supposedly realistic images, W. Eugene Smith may be more the rule than ihe exception among well-known documentary photographers (de Miguel, 2002). Even realist landscape photographers such as Ansel Adams (who railed against the subjectivities of "pictorialism") have adjusted the tone, contrast, and framing of their photographs to better express their own strongly held ideas about how the places they photographed "should look" (Brower, 1998). Other documentary photographers have done much the same in depicting culture and social life. In thinking through where the inventions of documentary image makers fit along what I have called the "empirical divide," there is much to learn from the work of social researchers themselves. Stimulated in part by the seminal work of Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1484 AMERICAN BEHAVIORAL SCIENTIST Becker (1986b) and Collier (1967), scholarly writing in this area has increased substantially in recent years through monographs (Banks, 2001; Biella, 1988; Emmison & Smith, 2000; Harper, 1982, 1987, 2002; Pink, 2001; Ruby, 2000), edited collections (Prosser, 1998), and an expanded array of journals (Visual Studies, The Journal of Visual Studies, The Journal of Visual Culture, Visual Anthropology, Visual Anthropology Review, etc.)- However, methodological treatments of image work within the social science literature are dominated by issues of research design to the neglect of personal vision and craft. Personal accounts by documentary photographers can alert us to other ways of thinking about empirical, visual inquiry. Dorothea Lange, for example, displayed prominently over her desk (he following quotation from Sir Francis Bacon, an early statesman for empirical inquiry: "The contemplation of things as they are, without substitution of imposture, without error or confusion, is in itself a nobler thing than a whole harvest of invention" (Lyons, 1966, p. 67). Lange arranged, cropped, sequenced, and edited her photographs to make documents that went beyond, in meaning and social impact, her camera's capacity to record the visible details she aimed it at (see Figure 4) (Coles, 1997). However, she also appears to have taken Bacon's statement seriously, at the very least as an alternative to the commercial work that occupied her attention prior to explicitly documentary projects and assignments. An empirical ideal for photographic purpose was also championed by Wright Morris, a documentary photographer, fiction writer, and essayist. Morris (1999) argued that "we should make the distinction, while il is still clear, between photographs that mirror the subject, and images that reveal the photographer. One is intrinsically photographic, the other is not" (p. 8). However, in what looks at first like a contradiction to the "mirror" ideal, Morris also noted, "Only fiction will accommodate the facts of life," adding that "our choice, in so far as we have one, is not between fact and fiction, but between good and bad fiction" (p. 103). Considered in light of his other writing, Morris's (1999) statement reveals what I have come to regard as a radical or root appreciation of empirical inquiry that is hard to find within the social science literature per sc. At the heart of this perspective arc two key ideas Morris developed more fully in both his photography and writing. First, every account of "the facts of life" will reflect some forms of inventiveness by investigators and reporters, not just in making photographs, or putting words on a page, or quantifying variables but also in linking observations of any sort to concepts, theories, or narratives—what Charles Ragin (1992) referred to as "casing." Second, depending on the intention, skill, and integrity of the investigator, these inventions can move an account closer to or farther away from "things as they are." Morris did not examine this provocative link between empirical inquiry and fiction in social scientific terms, but James Clifford (Clifford & Marcus, 1986) did just that a few decades later in characterizing "ethnography as fiction," but a kind of fiction that is not necessarily false or untrue (p. 6). In Clifford's Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 14X6 AMHRICAN ISH! IAVIORAL SCIKNTIST perspective, rhetorical "inventions" fall squarely within the tool kit of empirical social inquiry not as a substitute for detailed observation and systematic analysis but as their handmaiden. As Pink (2001), I larper (1998), and others have noted, this argument can apply as well to the "rhetoric" of photographic representation. The necessity of invention and rhetoric to productive empirical inquiry does not mean that "anything goes." However, it might be necessary to use artificial lighting to make a photograph that looks like what we can see in the field under "natural light" or to resequence raw film footage so that events and settings are more comprehensible and clear. To get comparable, empirically sound information, experienced field researchers recognize that they may need to alter a line of questioning from one informant to another. Along the same lines, it might be necessary to use different lenses, vantage points, or image-making strategies in one setting than in another. In some cases, a researcher might have to move objects around so that they can be better seen and recorded. There is also much to be learned about culture and social life from how participants respond to outsiders, including outsiders who come with cameras, videotape recorders, and questions that might otherwise never be asked (Biella, 1988). As Morris (1999) intimated, the choice is not between truth and invention but between inventions thai lead toward truths and those that lead away from them. This ties the soundness of empirical inquiry not only to techniques and methods but also to the ethics and integrity of the investigator. Though reflection and invention are not quite the same as objectivity and subjectivity, Coles (1997) spoke for both sets of terms in noting that to take stock of others is to call upon oneself—as a journalist, a writer, a photographer, or a doctor or a teacher. This mix of the objective and the subjective is a constant presence and, for many ofus, a constant challenge—what blend of the two is proper, and at what point shall we begin to cry "foul'.'" (p. 8) THREE EXEMPLARY PROJECTS A radical or root appreciation of empirical inquiry is hard to define beyond statements of principles such as those provided by Morris or Bacon; or critiques of scientism such as those offered by Marcus, Pink, and others; or a call to honesty and thoughtfulness such as that provided by Coles. It certainly does not turn neatly into a checklist of methodological dos and don'ts. And it falls far short of (or extends beyond, depending on your point of view) explicit guidelines Uncollecting or analyzing specific kinds of data—photographs or videotapes, interview transcripts, survey responses, or census tract figures. In the simplest terms, it calls for nothing more and nothing less than trying to ground ideas about the world as much as possible in observations of the world, to notice what is visible and account for it in ways that "get it right." Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Wagner/CONSTRUCTING CRI-DIUI.Ii IMAC.I-S 1487 Figure 5: The Namgay Family, Shinka, Bhutan, 1993. Photo copyright 2004 Peter Men/.el/menzelphoto.eom. Many social scientists spend their working lives trying to come as close as they reasonably can to this ideal. As illustrated by the three projects described below, so too do some documentary photographers. Although none of these projects has been embraced as bona fide "social research" by professional sociologists or anthropologists, each reflects a systematic approach to empirical inquiry, the intent to create new knowledge, and an effort to extend and refine social theory. After briefly describing these projects, I will turn to two related questions: First, how does the kind of empirical inquiry wc find in these three projects differ from what we have come to expect from social scientists1.' Second, what implications do these differences have for empirical, visual studies of culture and social life'.' MATERIAL WOULD Few documents provide a more provocative depiction of social and economic inequality than Peter Men/el's (1994) Material World: A Global Family Portrait, a survey in photographs, text, and statistics of the household possessions and routines of a single family from each of 30 countries (see Figure 5). In the six to eight pages allotted to each of these families, the authors present a wide range of data: a demographic profile and a paragraph or two about each country: Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1488 AMERICAN BEHAVIORAL SCIENTIST Figure 6: An Fnglish lesson in the school attended by 12-year-old Bangum Namgay, an hour's trek from her home in Shinka, Bhutan, 1993. Photo copyright 2004 Peter Menzel/menzelphoto.com. an array of captioned photographs showing the "daily life" of family members; a summary of each family's possessions and living space, including the "most valued possession" identified by different family members; and a brief account by the photographer. For each family we also arc provided with whal Men/el called Ihe "Big Picture," a single large photograph of family members standing or sealed among all their possessions, outside their home. These provocative images are interesting in their own right. They are rendered more informative by a legend that identifies objects and people and a list in the appendix of additional objects not included in the photo (Men/.el, 1994, p. 253), Both the photographs and text of Material World are clearly designed for impact, but pains were taken to make the impact empirically credible. The book provides a list of references and data sources and a table comparing all 30 countries on 22 different demographic variables. The selection of families is also described in enough detail, individually and in the aggregate, to alert readers to important qualifications and sampling questions and to provide some sense of Ihe immediate circumstances in which photographers worked. In his own account photographing the family portrayed in Figure 5, for example, Mcn/.cl (1994) wrote, For six days I lived with (he Namgay family in a twelve-house village an hour's walk from a 7-mile dirt road off a small paved road four hours from Thumphy, the capital. The Namgays had never seen a TV, an airplane, or for that matter a live American before and were as curious about me as I was about them. I had dinner Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Wagner/CONSTRUCTING CRKDIBI.i; IMAGHS 1489 Figure 7: The Skeen Family, Pearland, Texas, 1993. Photo copyright 2004 Peter Menzel/menzelphoto.com. with a different family every night, the same basic good food that I ate gladly with one hand as my legs ached from sitting cross-legged on the floor'. (My other hand fanned the flies from my food.) . . . Wild marijuana grows everywhere, but villagers feed it to their pigs after boiling it. The sounds were incredible: women singing in the fields as they harvested wheat, the murmur of monks chanting, the squeal of children playing, all without the haze of electronic noise I have unfortunately come to lake for granted. On the other hand, all was not paradisiacal. Animals and people excreted just outside the house and the family cooked inside on an open fire. (p. 78) We do not know from this comment alone exactly how Mcn/.cl decided what to photograph, but we do get some insight into the cultural contrasts and personal dispositions that shaped his image making in the field. A sympathetic reading of Material World requires that we temporarily ignore the cultural and economic diversity found within each country. However, Men/el (1994) presented the book not to challenge or discourage that kind of complexity but to resist another kind of simplification. As Men/el noted, "Newspaper, magazine and television stories almost always deal with the extremes: famine, flood, mass killing, and, of course, the life-styles of the rich and famous.... I wanted to give some insight into the rest of the world" (p. 255; see Figure 7). The empirical value of Material Wo rid rests in part on the study design and in part on an ability to elicit cooperation from the families themselves. This cooperation was inextricably tied to both data collection and reporting. Indeed, the Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. I4<)0 AMERICAN BEHAVIORAL SCIENTIST Figure 8: The Qampie Family, Soweto, South Africa. Photo copyright 2004 Peter Menzel/menzelphoto.com. power of the Material World accounts, family by family and country by country, hangs on making the visual comparisons and contrasts somewhat systematic. This applies with special force to the Big Picture. Inventories of household possessions have been described by anthropologists such as Collier (1967), Oscar Lewis (1965), and Janet Hoskins (1998). But they are given added punch by the technical virtuosity and documentary skill of the Material World photographers. As anyone who has tried it can attest, it is no small matter to arrange diverse materials so that they can all be seen at the same time let alone lo light and focus the array in ways that will produce a well-exposed and legible image (see Figures 5, 7, and 8). The same technical and representational skills that Material World photographers used to create empirically sound images could also be used lo misrepresent culture and social life. We do not know for a fact that they were not used in just that way, though we have many indications that this was not their intent. It is also clear that families willing to sit for such extended and intrusive "portraits" might differ somewhat from those who were not so inclined. And the idea of finding one family from each country flies in the face of more comprehensive and differentiated surveys. Though the imperfections of this research design are acknowledged rather than concealed, some readers might take them seriously enough to wholly dismiss what Material World has lo offer. However, a more appropriate lest of empirical merit is framed by the following two questions: Do we know more about social and economic inequality between different countries as a result of this book or less? And is what we know well grounded enough Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Wagner/CONSTRUC'I'lNt; ('Rlil)lIS!.]• IMACIIiS 1491 Figure 9: Two 15-year-old girls try on clothes in a dressing room, San Jose, California. Photo copyright 2002 Lauren Greenfield/VII. in empirical evidence to challenge speculation and ignorance? For some kinds of speculation and ignorance, I certainly think it is, GIRL CULTURE The questions noted above arc also worth considering in connection with Lauren Greenfield's (2002) documentary study Girl Culture. Like the creator's of Material World, Greenfield seemed intent on "getting it right" empirically— recording what she saw and what her participants had to say in ways that both document and raise questions about culture and social life. Indeed, (he artful juxtaposition of comments and images from different but related scenes is, in her hands, a tool of both personal and collective inquiry. In one cluster of photographs, for example, she recorded a range of women and girls working on their appearances in mirrors. Through another set of photographs, she documented a diverse array ol girls and women in di Herein forms of "dressing up" (sec Figures 9, 10, 1 Land 12). In putting together images of this sort, Greenfield (2002) has illustrated the exhibitionist dimension of feminine identity, a theme that plays back and forth between mass-market icons and personal appearance. As Greenfield pul it. The body has become the primary canvas on which girls express their identities, insecurities, ambitions and struggles. I have documented this phenomenon and at the same time explore how this canvas is marked by the values and semiotics of the surrounding culture, (p. 150) Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1492 AMERICAN BEHAVIORAL SCIENTIST Figure 10: Augusta, 22, the newly crowned Queen of the Cotton Ball, Chattanooga, Tennessee. Photo copyright 2002 Lauren Greenfield/VII. Figure 11: Exotic dancer Tammy Boom backstage at Little Darlings, Las Vegas, Nevada. Photo copyright 2002 Lauren Greenfield/VII. As an important variation on this theme, she also has reminded us that the exhibitionist equation works well only for a few women whose physiognomy matches well-advertised icons, and not even that well lor those. This encourages, as Greenfield sees it, the constant scrutiny and disaffection that women Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Wagner / CONSTRUCTING CRHniBI.H I MAGI'S I4l).i Figure 12: Elita, 6, at a birthday party where girls have their hair and makeup done, play dress up, model in a fashion show, and have a tea party, Hollywood, California. Photo copyright 2002 Lauren Greenfield/VII. express toward llieirown bodies and heightens the temptations of plastic surgery or physical self-abuse (see Figure 13). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1491 AMERICAN BEHAVIORAL SCIENTIST Figure 13: Erin, 24, is blind weighed at an eating-disorder clinic, Coconut Creek, Florida. She has asked to mount the scale backward so as not to see her weight gain. Photo copyright Lauren Greenfield/VII. In much Ihc same way dial Hiving GolTman (1963) called attention to the "total institution" as an ideal type that could characterize quite diverse organizations (prisons, monasteries, mental hospitals, boarding schools, etc.), Greenfield's (2002) work calls attention to "girl culture" as an ideal typical configuration of values, practices, and ideas through which women define and display their sexual identity. As Greenfield put it, "Understanding the dialectic between the extreme and the mainstream—the anorectic and the dieter, the stripper and the teenager who bares her midriff or wears a thong- is essential to understanding contemporary feminine identity" (p. 150). Like the authors of Material World, Greenfield (2002) has combined powerful photographs with other data, including extended interview comments by the participants in her study. In keeping with her intentions, these commentaries give her treatment of girl culture empirical depth and complexity: "As the photographs are my voice, the interviews give voice to the girls" (p. 152). The credibility of Greenfield's work is also enhanced by candor and caution in describing her own "inventiveness" and vision. She acknowledged that although the photographs "are about the girls 1 photographed ... they're also about me" (p. 152). At another point, her commentary reminds us that "infinite choices were made in the subject matter, the point of view, in the moment I depressed the shutter, in the editing. Ultimately, Girl Culture looks at a wide spectrum of girls through a very narrow prism" (p. 152). In another parallel to Material Culture, it is not just the photographs and interviews that create the "new knowledge" of Girl Culture but also the Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Wayner/CONSTRUCT! NC CRHDIBLH IMAGES 1-WS Figure 14: Wind gap pumping station, California Aqueduct, Kern County, 1985. Photo by Stephen Johnson. comparative framework within which Greenfield (2002) has placed them in this case, comparing women across age groups and social status instead of countries. Thoughtfully framed and sequenced, her photographs create a credible, multidimensional account, a kind of metaimage that both references and questions other images of women with which wc arc already familial'. THE CHEAT CENTRAL VALLEY The Great Central Valley: California's Heartland is a collaborative social history prepared by photographers Steven Johnson and Robert Dawson and the essayist and novelist Gerald Ilaslam (Johnson, Dawson. & Haslam, 1993). The book combines an extraordinary array of visual materials and a lengthy text that includes personal accounts and observations as well as a synthesis of scholarship from a wide range of disciplines—economics, agronomy, anthropology, and so on. These varied materials are organi/.ecl as convincing empirical evi deuce of the changing life and culture of the Central Valley of California. In the same chapter, we can find Farm Security Administration photographs from the 1930s, contemporary black-and-white photos made in the same geographical area (that look as if they could have been taken by Farm Security Administration photographers), contemporary color photographs of both old and new icons, other old photographs (some of which have been repholographed), satellite photographs, maps, and the reproduction of a landscape painting. Like the creators of Material World and Girl Culture, the authors of The Great Central Valley have described the process of their own creation, in this Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1496 AMERICAN BEHAVIORAL SCIENTIST Figure 15: Johnnie, Merced, California, 1975. Photo by Stephen Johnson. case through another book by Johnson (1993) titled Making a Digital Book. This companion volume provides additional details about how The Great Central Valley was designed and put together, both technically and conceptually. We learn that a prerelease version of Adobe Photoshop allowed Johnson to improve the legibility of old photographs by removing "cracks, serious scratches, and other artifacts of age," and that he also altered "contrast and brightness" to make some images more legible, but that the digital photo editing only went so far: "I was careful to respect the integrity of the original, however, and did not remove or add any real objects" (p. 21). Johnson's account of how ideas within the book came forth is equally explicit: Once I had settled on a basic grid [ for the design], my primary task was to find a relationship between the text and photographs that was integrated, but not directly illustrative. That really was the largest single design challenge, and the most time consuming. I had to know the photographs, read every word of the text, and imagine relationships, (p. 15) We might like to know more about the process by which Johnson "imagined" relationships between words and images in preparing The Great Central Valley, but the detail he has provided—including how he chose to present this study to Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Wagner / CONSTRUCTING CREDIBLE IMAGES 1497 Figure 16: Used cans, crop-dusting airstrip, Newman, California, 1984. Photo by Stephen Johnson. others—goes well beyond what we would expect from a social science research design. DOCUMENTARY AND SCIENTIFIC CONVENTIONS I have tried to make the ease, in an abbreviated way, that these three visual studies all provide credible, empirical accounts of culture and social life without necessarily respecting the conventions of contemporary social research. That is especially true for how the documentary image makers conducting these studies addressed the three challenges I noted earlier: creating empirically sound images of culture and social life, framing observations to highlight new knowledge, and challenging existing theory. Let us look at each of these in more detail. CREATING EMPIRICALLY SOUND IMAGES All three documentary studies make extensive use of recorded images to represent how culture and social life looks in particular times and places, and the images themselves provide a kind of information that is hard to represent in text alone. This is true for not only the sheer wealth of visual detail but also the Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1498 AMERICAN BEHAVIORAL SCIENTIST Figure 17: Discovery Bay, San Joaquin Delta, California, 1985. Photo by Stephen Johnson. precise imaging of physical and social environments from particular viewpoints, the juxtaposition of contrasting images, and the sequences and formats in which we encounter images as readers. Indeed, the photographs in these three studies go well beyond the common social science trope of "illustrating" ideas that are otherwise well accounted for in text. They provide instead a kind of content that is analytically interesting in its own right. In arguing for the credibility of this content, these documentary image makers give more attention to challenges of "recording" good evidence than do most social researchers. In Material World (Men/el, 1994), for example, we find not only a description of how the photographs were made in general but also individual accounts from photographers about each family they photographed. The two photographers working on The Great Central Valley (Johnson el af, 1993) offered their own accounts of what they were doing, photographically, in studying the Great Central Valley, as did Greenfield (2002) for her work with Girl Culture. The origins of these documentary studies are also described in terms that are more personal and situational than is typical for social science study designs. Greenfield (2002) noted that she was "enmeshed in girl culture before I was a photographer, and I was photographing girl culture before I realized I was working on Girl Culture" (p. 149). Johnson (1993) reported that he "embarked on the Central Valley project to better understand the place that made me a landscape photographer" (p. 43). Menzel's (1994) account of what led him to the kind of Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Wagner/ CONSTRUCTING CREDIBLE IMAGES 1499 data reported in Material World refers to not only the 1994 United Nation's International Year of the Family and his previous work as a photo journalist but also a program he heard on the radio about marketing a sex-fantasy book by the pop star Madonna: "The book and the singer seemed to hold more interest for people than the pressing issues of our day. I thought the world needed a reality check" (p. 255). Evidence about how individual images were made and the personal interests of the researcher does not in itself make investigations more empirically sound—see Biclla (1988) and Ruby (1976) for contrasting views on this. However, it can help us determine how close a study comes to hilling its empirical marks. A research design can help in that regard as well, of course, and thai suggests the value of providing both, along with some sense of where the author thinks the findings of a study might or might not apply. For example, Greenfield (2002) has commented that "Girl Culture is my photographic examination of an aspect of our culture that leaves few women untouched" (p. 150), but she also has noted that her book "does not attempt to represent the experience of all girls in America, or even the full and rich experience of any girl I photographed" (p. 150). HIGHLIGHTING NEW KNOWLEDGE Lots of information is available to people that they do not care about, or think they already have, or reject, but new information can sometimes add a new dimension to how we think about things. The uncertainties of this process present a real challenge for people who want to undertake empirical social inquiry. Who will care about the particular inquiry they have in mind? And what forms arc available for presenting new knowledge to those particular people'.' The answers social researchers develop to these questions almost always involve publishing articles and books for specialized academic communities and markets. As they see it, for knowledge to be really "new," it has to be new for colleagues already hard at work studying related questions and phenomena. Documentary image makers approach this challenge somewhat differently. They are not particularly interested in creating knowledge that appears to be new only to small groups of social scientists. Like social researchers, they would like their work recognized and respected by professional peers. But documentary image makers also pilch their inquiries to other audiences, including the participants in their study and other people like them, and to members of the public who may harbor ideas about the visual evidence they have put together. As one step toward reaching this broader audience, some documenlarians (including those I have described here) represent their work as the outcome of a personal journey that led to new insights and understanding. Johnson (1993) noted that although making photographs for The Great Central Valley began on familiar footing, it "grew into the discovery of a place T didn't know very well. It became an exploration of land use, water use, agricultural practices, racism and Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1500 AMERICAN BEHAVIORAL SCIENTIST poverty" (p. 43). In establishing points of personal connection with both professional and public audiences, Greenfield (2002) reported thai "Girl Culture has been my journey as a photographer, as an observer of culture, as part of the media, as a media critic, as a woman, as a girl" (p. 149). This personal and public rhetoric contrasts with how social researchers index their own reports to specific research publications and communities (Richardson, 1991). The documentary image makers reviewed here also have given much more attention than social researchers usually do to issues of editing, layout, and visual representation. Not only have they made explicit the aesthetic dimensions of this work but they also link design issues directly to both analysis and audience. I have already noted Johnson's (1993) extended account of what it took lo prepare The Great Central Valley in book form, but Menzel (1994) and Greenfield (2002) also offered explicit commentary about designing their books. As another illustration of this emphasis, Greenfield has distinguished her contributions to Girl Culture from other instances in which the same photographs appeared for other purposes: "While 1 often can't control the picture editing, writing and design in my work for magazines, the selection and presentation of photographs in this book are my own" (p. 152). For all three documentary image makers and authors, the boundary between research participants and public communities is also blurred, all the more so because each has encouraged distribution of this work in other forms. The Great Central Valley was at first a documentary project, then an exhibit in the Central Valley itself and a symposium, then a book, and later a book about the book. Girl Culture also began as a documentary project, elements of which appeared in mass-market publications, and Ihe book is now complemented by a traveling photo exhibit and a Web site that includes an on-line photo gallery, transcripts of all 20 interviews reported on in the book, a teaching guide, links to organizations working on related issues, video interviews with Greenfield, and an opportunity to participate in related on-line forums. The work brought together in Material World has also appeared in other publications, and a CD-ROM is now available that both replicates and extends the content of the book. Taken together, these diverse activities and media provide a larger and more variegated public presence than we would expect from a publication alone, let alone a publication addressed primarily to social scientists. The new knowledge available lo research participants and the public through these documentary materials is available lo sociologists and anthropologists as well, but it is not inscribed in mainstream social science journals. Because these documentary image makers have not framed their work in terms of social research per se, it is nol clear where it would lit to best advantage in those venues. But Ihe rhetorical conventions of published social research—the emphasis on words and numbers, accompanied at times by figures and charts, organized around arguments and summarized "findings"—is also problematic in its own terms for documentary image makers. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Wagner / CONSTRUCTING CREDIBLE IMAGES 1501 These problems become apparent when we try to imagine converting any of these three studies into standard social science reports. An abstract or synopsis of each might be noteworthy but would also fall far short of the new knowledge we are likely to acquire from reading each work as a whole, some of which occurs as a process of elicited meaning and inquiry. As Paul Kennedy noted in his introduction to Material World, "The real benefit to learning that the reader can extract from this project depends on going into the details, especially on a comparative basis. New kinds of valuable inquiry can be generated by such detailed observation" (i.e., "observing" the book itself; Menzcl, 1994, p. 7). As a related point of contrast, the balance between evidence and interpretation in the documentary projects reviewed here is weighted more toward evidence than is customary for social science research reports. That may make documentary studies somewhat more ambiguous than social scientific reports, but it does not make them any less empirical. CHALLENGING SOCIAL THKORY Girl Culture includes an introduction by Joan Jacobs Brumberg, a professor of human development and women's studies at Cornell University, and Greenfield's (2002) own commentary refers to a few scholarly studies that helped shape her thoughts. The Great Central Valley is heavily referenced to the work of historians, geographers, and policy analysts. And Material World lists numerous sources that someone could consult to learn more about the countries and issues it examines. However, just as none of these projects take social researchers per se as their primary audience, neither do they frame insights to readers as a contribution to academic scholarship. Johnson (1993) is quite explicit about his interest in avoiding both romantic and academic genres: "None of us wanted this project to become another photography book idealizing a landscape," he noted, "nor did we want the book to become an historical dissertation" (p. 15). This apparent neglect of disciplinary scholars goes hand in hand with the interest of documentary image makers in attracting other audiences. However, it also reflects alternative ideas about where social theories are most likely to be found, acquired, and contested. The "theories" that social scientists pay the most attention to are inscribed explicitly in published social science texts. Docu-mentarians might acknowledge this kind of theory as well. But they also attend to a wide range of cultural materials in which social theories are more embedded than explicated—texts, of course, but also news accounts, folklore, and mass-media imagery. Instead of contested theories and hypotheses, the documentary projects I have described here are designed to highlight contrasting ideas and imagery. These image-based challenges to social theory can mirror exchanges among academics about different theoretical perspectives, interpretations, and data Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1502 AMERICAN BEHAVIORAL SCIENTIST sets, but they can take other forms as well. For example, Johnson el al. (1993) reproduced in The Great Central Valley some policy documents and photographs that they then called into question through juxtapositions with other documents, their own photographs, or through the testimony of local participants. Greenfield (2002) both photographed and critiqued some of the images that the people she studied respond to in constructing their identities. With admirable candor, she also noted that as ajournalist, she helped make some of those images herself. Menzel (1994) saw Material World as a way to not only illustrate 'The great differences in material goods and circumstances that make rich and poor societies" (p. 7) but also challenge less credible ideas, some of them supported by images he had helped create through previous photographic assignments. In each case, the documentary photographs presented by these authors are framed to challenge other images that reflect existing, largely implicit, and widely held ideas about culture and social life—elements of social theory, by any other name. EMPIRICAL INQUIRY AND PUBLIC DISCOURSE In terms of empirical social inquiry, the documentary studies I have described above arc exemplary. Other studies that are called documentary simply because they include "realistic" photographs of people and places may actually fall more appropriately on the other side of the empirical divide, where speculation, projection, fantasy, and introspection hold sway. We can approach work on either side of that divide without prejudice, and projects that straddle it can be stimulating and provocative. However, in looking to documentary image making for empirically sound accounts of culture and social life, I suggest we seek works—such as those reviewed here—that not only "look interesting" but also reflect a clear commitment to empirical inquiry. Having said that, if we think of Girl Culture, The Great Central Valley, and Material World only as documentary work, we isolate what we can learn about empirical inquiry through projects of this sort from how wc think about social research. A more productive strategy is to consider each project as an instance of empirical social inquiry, analytically defined. Instead of asking, What is the difference between documentary photography, narrative accounts, and sociology or anthropology? we might ask, How does empirical social inquiry look when practiced by skilled sociologists or anthropologists, and how does it look when practiced by skilled documentary photographers, journalists, and essayists? As a partial answer to these questions, I have summarized in Table 1 some of the contrasts noted above between social science and documentary studies. These contrasts suggest that in some circumstances, one approach to empirical social inquiry might work better than the other. For example, if we want to build a written literature around a distinctive set of concepts and questions—a disciplinary tradition, so to speak—the conventions of social science have the most to Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Wagner / CONSTRUCTING CRKDIBI.H IMAGES 1503 TABLE 1: Two Modes of Empirical Social Inquiry Social Science Documentary Study Purpose Research design Data sources Data collection Data analysis Reports and representations Audiences Framing new knowledge Theory building Develop new knowledge and understanding of culture and social life through empirical investigation and scholarly "works." Dedication to explicit research design, including a priori rationale for linking questions with appropriate data sources, scope of data desired, identification of analysis strategies, and so forth; emphasizes testing ideas through narrowly bounded inquiry; separation of personal interest from logic of inquiry. Direct observation, field recording of images and interviews, found artifacts, and so forth plus surveys, analysis of aggregate data sets, and so forth. Emphasis on getting enough data points to meet requirements spelled out in research design; larger sample size preferred to more detailed observation of particulars. Precursor to reporting and representation; systematic use of discrete analysis strategies; analysis restricted to bounded data sets. Representation as afterthought to data analysis; focus on matching reports to publication options; primacy of summary, report, and argument. Specialized research community as primary audience but passing interest in public and popular constituencies, including policy makers. New knowledge as extension, complement, or alternative to existing and explicit social theory. Emphasis on propositions inscribed in the social science research literature: "competing propositions" as primary content drama. Develop new knowledge and understanding of culture and social life through empirical investigation and public "works." Casual attention to research design; implicit and diffuse statement of research questions, data sources, and so forth; emergent rather than a priori focus and questions; emphasizes exploring, investigating, and examining phenomenon, place, people, or idea through broadly bounded inquiry; integration of personal interest and logic of inquiry. Direct observation, field recording of images and interviews, found artifacts, and so forth. Explicit attention to recording challenges and media; interest in presentation of quality documents and data sources. More detailed observations preferred to larger sample size. Closely integrated with issue of representation; push toward coherence and clarity through multiple analysis strategics; unrestricted data sources. Great attention to issues of representation, aesthetic ideals/principles; reports designed for power and effect; primacy of narrative, example, and collage. Public and popular constituencies as primary audience but passing interest in specialized communities (policy makers, researchers, research participants, etc.). New knowledge as images, concepts, perspectives that are new to public or to targeted communities. Emphasis on ideas and principles embedded in public media and discourse: "contrasting images" as primary content drama. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1504 AMERICAN BEHAVIORAL SCIENTIST offer. Why? Because they require that new work be tightly indexed to the work of other scholars who have wrestled in writing with similar questions and concepts, a kind of intertextuality that both reflects and stimulates the evolution of a literate community. But if we care less about literature building than about communicating with diverse constituents, documentary work with images has real advantages of its own. These advantages certainly apply to the challenge of informing public discourse, but they also have special relevance for human service professionals working in diverse communities. Teachers, for example, as well as social workers, community organizers, and health care professionals, may find documentary conventions more agreeable and productive than social scientific approaches in studying local clients and communities. They do not need to know if new ground is being broken for the disciplines of psychology, sociology, or anthropology to learn something valuable from looking at videotapes of student small group discussions, or from making and examining photographs of institutional events and routines, or from working with young and old community members to mount an exhibition of photographs and stories that document neighborhood and family traditions. Reading something thoughtful can help teachers and other human service professionals think about this kind of documentary inquiry. But community members can conduct thoughtful, empirically sound, and useful studies of local culture and social life without having first reviewed the relevant social science literature. The corollary also holds true: Empirical studies of culture and social life can generate valuable insights for local practitioners and clients without breaking new ground for professional social scientists. (As a related matter, if we want practitioner and community perspectives to become more visible within a research literature, we will have to do something more than support individuals in studying the culture and social life of their own classrooms, schools, workplaces, and local communities—however valuable that might be to the people involved.) The contrasting merits of documentary study and social research as resources for field-based professions extend as well to undergraduate curricula and students. Although social scientific knowledge is essential to an informed citizenry, so too are documentary studies and the ability to think clearly about credible images of culture and social life. And it is naive to think that students can learn to assess these matters thoughtfully without trying to construct credible images of their own. Whether opportunities for that kind of experience fall under the rubric of visual studies, documentary studies, field research, liberal arts, cultural studies, media studies, or social science may matter little. But engaging students in producing and questioning the kinds of documentary studies 1 have examined here—and struggling with related questions about evidence, representation, audience, imagery, and ethics—represents a good investment in young people and civic culture, a better investment, perhaps, than the Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Wagner/CONSTRUCTING CRIiDIBLIi IMAGHS 1505 kind of disciplinary specialization that stands in currently for a liberal arts education. Some sociologists and anthropologists have put recorded images to extremely good use within their own research and teaching, and they have a lot to say about how to approach these issues ethically and with good sense. 1 low-ever, and understandably so, their work is directed typically toward advancing the field, educating colleagues, and orienting new members to their own discipline or profession. It is quite another matter to educate citizens about the complexities of visual representation and the promise of visual studies or to acquaint them with what it takes for photographs, films, and videotapes to provide empirically sound accounts of culture and social life. Being smart about that requires that we learn what we can from not only social scientists but also documentary photographers and filmmakers, at least some of whom celebrate both art and empiricism and aim for both telling images and telling truths. REFERENCES Agcc, .1., & livans, W. (1960). Let us now praise famous men. Boston: Houghton Mi lit in. (Original work published 1939) Banks. M. (2001). Visual methods in social research. Thousand Oaks. OA: Sage. Becker, 11. (1986a). Do photographs tell the truth? In Doing things together: Selected papers (pp. 273-292). livanslon. II.: Northwestern University Press. Becker, 1-1. (1986b). Photography and sociology. In Doing things together: Selected papers (pp. 22 1 -272). Hvanston, II,: Northwestern University Press. Biella, P. (1988). Against reduclionisni and idealist scli-rcl'lexivity: 'The llparakuyo Maasai T'ilni Project. In J. R. Rollwagen (lid.). Anthropological'filmmaking [pp. 47-72). New York: I larwood. Brower, K. (1998). Photography in the age of falsification. Atlantic Monthly, 2