Contrasting Images, Complementary Trajectories:
Sociology, Visual Sociology and Visual Research

Jon Wagner
Division of Education
University of California, Davis

Published as:  Visual Studies 17 (2), pp. 160-171, 2002
[reprinted from  Sociological Imagination 38 (1/2) pp7-27, 2001]

A little over 20 years ago I had the pleasure and good fortune to edit Images of Information, a volume of essays about the use of still photography in the social sciences (Wagner 1979).  In that book, I and a dozen contributors raised questions about where photographs could productively fit within the epistemological and reporting conventions of the social sciences.  However, our primary emphasis was less theoretical than rhetorical and pragmatic.  We argued in favor of giving greater attention to photographs and other forms of visual representation in social science research and teaching, and we described cases and strategies that exemplified how this might be done.

Since Images of Information appeared, numerous new examples of the kind of work we had in mind have seen the light of day.  Some of these have fallen clearly within a loose constellation of work that might be called "sociology itself." Related work has appeared in anthropology and through institutional or professional applications for which sociologists and other social scientists have provided intellectual leadership.  New work reflecting similar themes and perspectives has also appeared along seams between the social sciences and the professions, the arts, journalism, and the humanities. 

As a complement to this increasing array of visually-oriented studies, two books have recently appeared that take a stab at articulating the array as a whole with the work of sociologists and other social scientists: Jon Prosser's Image-based Research (1998) and Michael Emmison and Philip Smith's Researching the Visual (2000).   Given the lack of such volumes during the intervening 20 years, these recent books invite a reassessment of the arguments and examples embedded in Images of Information.  In what ways, if any, has social science discourse about photographs, imagery and the visual changed since 1980?  What does this suggest about the visual prospect of sociology itself?  And what directions can we identify now that seem more or less productive for the next decade or two of social and visual research?

The books by Prosser and by Emmison and Smith provide intriguing but partial answers to these questions, and I'll look closely at each with just that in mind.  Before I do, however, let me review a few changes that have taken place over the past two decades in sociology and the social sciences.  These changes help define the distinctive context in which Images of Information first appeared.  More importantly, they reveal trajectories that have not only brought us Image-based Research and Researching the Visual but also implicate the prospects for a sociology in which images and the visible are more routinely affirmed. 

What's changed in sociology?

When I and my colleagues were putting together Images of Information we were optimistic about the future of image-based work within sociology itself.  Our optimism was fueled by Howard Becker's (1974) thoughtful account of the correspondence between sociology and photography. Those of us excited about this correspondence explored our ancestral roots -- the early use of photographs in the American Journal of Sociology (Stasz 1979), the consultation that sociologists Robert and Helen Lynd provided to Roy Stryker in shaping shooting scripts for the Farm Security Administration photo-documentation project (Fleischhauer and Brannan 1988), and the bridge that some documentary photographers such as Louis Hine and Paul Strand had forged between academic training in the social sciences and field work with the camera.

Drawing on these antecedents and exemplars, "visual sociology" made sense to us.  We might not always agree about what this meant, but we envisioned sociology as a discipline that had room for working with photographs and other images.  The sociologist Bruce Jackson was doing just this kind of work in his studies of prison life (1977).  William H. Whyte was taking pictures of pedestrians in Manhattan public spaces, accounts of which appeared in both the scholarly journals and the New York Times Magazine (1972).  Photographs also appeared in Kai Erikson's study of the Buffalo Creek Flood (1976) a book that won the 1977 Sorokin Award of the American Sociological Association. 

In addition to these promising indicators within the discipline itself, sociologists and other social scientists working in the allied professions of education, architecture and medicine were collecting and representing data about the visual dimensions of social life.  Documentary photographers and filmmakers in the 1970's were also adding to what we knew about particular social groups, settings, and locales -- though, as Becker had noted in his 1974 essay, with less theoretical foresight that we might like as sociologists.  Looking beyond still photography we could find parallel social inquiries making use of film and video tape.  John Collier (1967) had both pioneered some of this work and written about it in terms that applied as well to sociology and education as to his home discipline of anthropology.  Indeed, in the years just before Images of Information was published, the sociologist Hugh Mehan was video-taping school classrooms to conduct the micro-ethnography reported in Learning Lessons (1979). 

All this by way of noting that in the late 1970's, the possibilities for putting photographs, film and video tape to good use within sociology and within other social science disciplines seemed very promising indeed.  And they were.  And they still are.  But sociology and the social sciences are not quite what they were back then, nor is clear what falls inside or outside the disciplines.

Methodological Communities

One way in which the disciplines have changed is reflected in the growth of scholarly communities defined as much by research method as by theoretical perspectives or objects of inquiry.  New methods-oriented journals, professional associations, conferences, and courses of study have emerged that cut across what were seen formerly as the core social science disciplines; others that were formerly tied quite closely to a particular social science have broadened their scope and appeal.  This kind of methodological confluence occurred in the 1950's, 60's and 70's among social researchers working with numbers and experiments.  "Survey research" became a cross-disciplinary method, one that sociologists, anthropologists, psychologists and political scientists could discuss within their own disciplines and also with each other.  ANOVA, "Likert scales," and two-tailed T-tests became elements of a cross-disciplinary research lexicon, and scholars conversed with each other about such matters accordingly.

Beginning some time earlier, but accelerating since the early 1980's, similar forms of affiliation emerged around research methods involving interviews, life histories, and field observations.  And, in the same way that quantitatively or experimentally orientated researchers talked with each other about different kinds of numerical data (e.g., parametric and non-parametric measures) social researchers who did field work, interviews, and the like began talking and writing about the kinds of data linked closely to their own approaches -- words, for example, artifacts, physical traces, or images.  In support of these cross-disciplinary examinations of field work, new journals, conferences and networks emerged around issues of "qualitative research."  William Filstead (1970) had edited a collection of essays under this title in 1970, but the full force of this orienting term did not appear until the 1980's and 1990's.  As it did, some existing research forums also broadened their focus and readership.  The Journal of Urban Ethnography became Contemporary Ethnography.  More recently, Cultural Anthropology Methods was reconstituted, with the same editor, as a new journal called Field Methods.  And in the coming year or so, Visual Sociology will shift to a new publisher and the new title of Visual Studies.

Broadened discourse about field methods brought scholars from different disciplines to new insights, not only about methods, but about theory and reporting conventions.  For example, while the rhetoric of written sociology has been critically examined throughout the discipline's history, these examinations rarely captured the attention of mainstream scholars.  But that changed in the 1980's, as anthropologists, linguists, folklorists, cultural historians and literary critics joined in common cause to look at issues of "representation" ("Representations" was the title of yet another new journal emerging at this time).  One by-product of this attention to written representations was an increasing understanding of the implicit theories that sociologists and other social scientists embedded in their writing, knowingly or not, through metaphor, synecdoche and both literal and figurative "imagery" (Becker 1986; Clifford and Marcus 1986; Gusfield 1981; Hunter 1990; Van Maanen 1992).

Many distinctions between the social sciences and their sub-disciplines survived these cross-disciplinary inquiries into field methods and representational strategies, but not all of them did.  Indeed, new specialties emerged within disciplines that, while peripheral to the discipline itself, became hot beds of cross-disciplinary inquiry.  Media and communication studies emerged as such a specialty, and so did gender studies and studies of material culture.  As we can see in the two books under review, visual sociology was yet another, though not always by that name.

Policy research, occupations and academic programs

Among the redefining forces shaping sociology and the other social science disciplines over the past twenty years, social policy has played an increasingly influential role.  As a manifestation of what Raymond Mack called, "fiduciary drift"-- i.e. the role that research funding plays in shaping research questions -- scholars have used their disciplinary tools to conduct policy-related research about the mass media, children's welfare, aging, and changing family structures.  State and foundation funded research of this sort draws scholars into cross-disciplinary affiliations that complement those that I noted above regarding research methods.  Some disciplinary purists have resisted these redefining influences, either quietly or loudly, and it's important to remember that hand-in-glove affiliations between research scientists and the state (or private funding sources) were not always routine (Kerr 1962).  However, they are now taken for granted by most social scientists, and the ability to secure research funding of this sort is affirmed in faculty review criteria at most research universities.

Over the past 20 years, new occupations also opened up in response to a changing economy and changing demographics, and training programs emerged to prepare people for them.  Undergraduate students voted with their feet to redefine coursework and programs in response to their own career ambitions and intellectual interests.  Many sociology programs developed applied strands, formed closer links with allied professions, or taught methods courses that students could put to good use in a wide range of occupations.  But there were limits to how much existing departments could accommodate the multiple ambitions and interests of new generations of students, and so new programs also emerged and expanded -- not only in computer science, but in communications, media studies, ethnic studies and cultural studies.

These changes in how the social sciences were presented to undergraduates also worked their way through graduate and professional schools.  In education, as one notable example, the dominance of psychology and philosophy as orienting disciplines gave ground to cross-disciplinary perspectives on cognitive science and the socio-cultural contexts of schooling.  These new orientations embraced and required new research strategies and designs.  New methods -- frequently qualitative in part -- were required to study not just how well students did on tests and what they were supposed to know, but what they actually knew, what they thought they knew, and the circumstances within which they might or might not learn and demonstrate such things.  In response to these shifts in theory and in the methods of educational research, words and images -- photographs, drawings, cognitive "maps" and video tape -- appeared as increasingly valuable forms of data, a development that had parallels in medicine, business and other applied fields.

Interdisciplinary careers

An often-overlooked stimulus to the changes I've noted above was the somewhat missionary extension of research sociologists from the late 1960's through the 1980's throughout a wide range of institutional settings.  From the 1950's through the 60's and into the early 70's higher education expanded rapidly, the social sciences in particular.  Senior scholars were tapped to head new departments and prepare new cohorts of Ph.D. students.  Junior scholars were bumped up quickly into positions of departmental responsibility.  New doctoral degree programs were created and produced an increasing number of social researchers.

When the boom stopped, it took awhile for programs to adjust their output of research Ph.D.'s.  I can well remember the consoling words of my faculty mentors in 1971, the year I got my own doctorate in sociology at the University of Chicago, that this was the worst job placement year for sociologists that anyone could remember.  They quickly added that next year would have to be better because the dearth of sociology positions at research universities certainly could not get any worse.  And yet it did, and continued doing so for almost 15 years.

One consequence of producing more new doctorates than there were research faculty positions was that research sociologists, anthropologists, and psychologists took positions outside the research university itself, in small colleges, state colleges, and junior colleges.  Recent Ph.D.'s also migrated into business or industry, or into government and applied fields where the perspective and research methods of the scholar were valued more than the scholar's discipline per se. 

Twenty years after Images of Information appeared, these changes present us with a world in which new intellectual communities and networks have emerged around methods of social inquiry that involve texts and image, within which text and images have increasingly become objects of investigation in their own right.  A new generation of scholars has emerged whose academic training has been shaped by "cross-disciplinary" programs in ethnic studies, women's studies, cultural studies, media studies or the professions, some of whom still claim sociology or another traditional social science disciplines as their intellectual home.  We also find sociologists, by training or self-identification, conducting research from positions both inside and outside traditional departments of sociology -- sometimes through applied work in a variety of institutional settings and sometimes through their own undergraduate or graduate teaching assignments.  As a result, the research colleagues of sociologists might sometimes be other sociologists, but in many cases they are students, agency or program staff, school teachers, or scholars and practitioners from other fields.

Where does image-work fit within this new constellation of disciplines, sub-disciplines, methods, programs, situations and subjects?  Are images per se a proper object of inquiry for sociologists  -- something analogous to the family, interpersonal relationships, religion, cities, or schools?  Or is it more appropriate to treat them as an important, but hardly defining, dimension of traditional subject areas?  What about the visible social world, or imagery of the social order?  Are these at the heart of sociological inquiry, or are they footnotes, appendices, oddities? 

"Image-based Research"

An engaged reader doesn't need to know the details of the changes I've reviewed above to profit from reading Jon Prosser's edited collection, Image-based Research, but it might help in a couple of respects, one of which is the juxtaposition that appears in this volume between interdisciplinary and disciplinary arguments. 

In the first of seven chapters within Part I, the "theoretical overview" section, Marcus Banks describes the transformation occurring in visual anthropology from an earlier and almost exclusive emphasis on ethnographic film to a much broader array of visual representations and visible phenomena.  In an extremely thoughtful essay, Banks captures the tension between expanding the range of visual "objects" while at the same time narrowing -- or preserving -- disciplinary tenets of anthropological work.  As Banks puts it:

"Visual anthropology is coming to be understood as the study of visible cultural forms, regardless of who produced them or why.  In one sense this throws open the floodgates -- visual anthropologists are those who create film, photography, maps, drawings, diagrams and those who study film, photography, cinema, television, the plastic arts -- and could threaten to swamp the (sub)discipline.  But there are constraints. . . [for] the study of visible cultural forms is only visual anthropology if it is informed by the concerns and understanding of anthropology more generally," (p. 11).

Following this splendid opening essay by Banks, Doug Harper makes the case for continued and expanded work in visual sociology within and beyond the genre of the "realist tale."  In making this argument, Harper acknowledges explicitly the tensions that have appeared between empirically-oriented sociologists and their semiotician and cultural studies colleagues.  However, in contrast to Banks' attention to a wide range of "visible cultural forms," Harper focuses primarily on the photograph and the research photographer.  While he sees great and continuing promise for visual ethnographies, Harper also acknowledges the postmodern critique of the "documentary" gaze as a form cultural domination.   "Visual sociology must confront this critique," he notes.  "Some will, of course, take it seriously enough to never photograph in the field again.  Others will continue to form creative solutions which address at least some of the critique" (p. 36).

These two essays by Banks and Harper each push visual studies towards broader subject matter and more diverse relationships with research subjects, but they also re-affirm the intellectual and theoretical centrality of their respective disciplines -- anthropology and sociology.  In the third chapter, Paul Henley arrives at a somewhat different position by examining the role filmmaking can play in field work, not only as a medium for illustrating research findings, but as an intervention for generating new ethnographic insights.  He pegs some of this potential to opportunities created by new visual recording and presentation technologies, some of which (including synch-sound video cameras and CD-ROMs) can reconfigure how images and text contribute to anthropological theorizing. 

By the time we arrive at Chapter Four -- a thoughtful essay by Brian Winston about the partiality of photographic evidence -- the kind of disciplinary integrity that Banks and Harper affirmed is coming a bit undone.  The disciplinary paradigm gets another small boost in Chapter Five -- Orla Cronin's examination of the meaning and psychological significance of family photographs -- but the psychology Cronin refers to here is pretty broad; it encompasses not only clinical and research contexts but individual and collective meaning making as well. 

Howard Becker adds his own problematizing commentary in Chapter Six by noting that, "A photograph taken by a 'visual sociologist,' 'documentary photographer,' or 'photo journalist' may be very similar.  However the significance and legitimation of that photograph is not to be found in attaching any particular 'catch-all' term but in the response it generates in those who perceive it"  (p. 84).  The framework that Becker recommends for thinking these matters through presents the academic disciplines themselves as epi-phenomena of a more general relationship between social organization, audience, work, and work products.  He concludes by noting that, "for sociologists and other social scientists, these examples provide a warning against methodological purism, an illustration of the contextual nature of all efforts to understand social life." 

With Becker's framework fresh in mind the chapter that follows -- Jon Prosser's commentary on the status of image-based research -- appears as a new kind of community study.  In this case the focus is not the social science disciplines per se, but the "qualitative research" community, and, within that community, a small group of scholars who undertake "image-based research."  Prosser both laments and examines what he sees as the "marginalization" of image-based research, and his commentary is both thoughtful and trenchant.  But the center of Prosser's firmament is at the margins of those reflected in the work of Banks and Harper, while the disciplines at the center of theirs --  anthropology and sociology -- are at the margins of Prosser's own, interdisciplinary conception of "qualitative research."

I have described these first seven chapters in such detail not only to alert prospective readers to the diverse content of Image-based Research, but also to illuminate how this well-crafted volume not only reports on, but exemplifies tensions between the disciplinary past of visual sociology (and visual anthropology) and what appears to be a somewhat interdisciplinary present and future. 

Within the diverse contributions to this volume, some themes do appear more than once, and one of these is the attention given to photographic imagery itself.  With the exception of the Banks and Winston chapters, the kinds of images referred to in the first seven chapters are usually photographs.  In the remaining chapters of this book, photographs continue to appear as a primary medium for conducting the kind of "image-based research" Prosser is calling to our attention.  Part Two contains five chapters that explore the role of "Images in the Research Process," all of which refer to photographs, some of them exclusively.  But this section also contains an important exception, a critique Holida Wakefield and Ralph Underwager of the "application of images in child abuse investigations." 

Two things make the Wakefield and Underwager chapter an exceptional read:  First, the authors do not identify themselves primarily with any particular academic discipline, but as scholars -- and policy analysts -- with expertise in a specific strand of interdisciplinary research about the "child witness."  Second, their analysis focuses on the full range of visual representations used to interview children in child abuse investigations -- drawings, dolls, books, puppets, and photographs.  Wakefield and Underwager consider each form of visual representation on its own and relative to the other forms and to the merits of sensitive and insensitive verbal interviews. Precisely because their analysis is neither disciplinary nor photo-centric (and because it is about something other than images per se) it provides what is perhaps the most provocative assessment of the situated interplay between images, objects, speech and text that I've read to date.

Part Three of Prosser's book provides case studies and examples of image-based research in practice.  Of the six chapters that appear in this section, three refer to photographic inquiries situated in schools, classrooms and families.  Among the others are Terry Warbuton's intriguing examination of how teachers are represented in cartoons, and Noreen Wetton and Jennifer McWhirter's report on the use of a "draw and write" strategy to elicit children's knowledge of health and health services.  In the final chapter of the book, Elizabeth Chaplin examines the life and work of John Constable as a case study that has less to do with the artist as an individual actor than with how the "meaning" of both artist and artworks are made and remade through collective action, mediated struggle, and the marketplace.

Each of these last six chapters examines interesting facets of image-based research while also being "about" something else:  how teachers are represented by children and in the mass media, how family members communicate with each other, how people think about health issues, the production of art works, and so on.  By pointing in both directions, towards images as subjects and towards images as data about other subjects, some of these latter chapters in Image-based Research help clarify the value of visual research within different domains of inquiry.  These domains may not correspond closely to the academic disciplines per se, but neither do they float free of external referents for the images they bring to our attention. 

Another theme running through these varied accounts is the connection several authors make between image-based research and the tenor and scope of researcher-subject relationships.  Several contributors to Image-based research are well aware of the complex challenges of doing field work in natural settings and express concerns for the welfare of research subjects as well as for the researcher's responsibility to make valid observations and link them systematically to defensible explications of cultural and social life. "Reflexivity,"  a key concept for understanding these challenges, is examined explicitly in four of the chapters.  Prosser's own formulation of these issues is particularly cogent.  He writes,

Validity in sociological and anthropological field studies has been focused essentially on methods rather than on people and human processes. . . This construct and application of validity is now seen by many to be limited if not inadequate for human enquiry, particularly studies of culture which are better supported by notions of validity based on an interactive, dialectical, collaborative logic (p. 104). 

As a replacement for a traditional emphasis on non-interactive notions of validity, Prosser recommends a concept of "contextual validity," which he describes in the following terms:

Judgments and claims of contextual validity are best made essentially via reflexive accounts but also through representation.  Reflexive accounts attempt to render explicit the process by which data and findings were produced.  Representation for image-based researchers reflects not only the sources of information in terms of pictorial codes but also the mode of communicating findings to recipients of research. 

Attending to the social contexts in which field work is conducted and reported positions the researcher as a person who works closely with other people -- both research subjects and research audiences.  From this position, photographs and other images can be collected or made, or used as interview prompts, and several chapters in Image-based research examine one or more of these field research strategies.  As a complementary orientation, several other chapters -- including those by Rob Walker and Ron Lewis, Richard Chalfen, Michael Schratz and Ulrike Steiner-Loffler, and Noreen Wetton and Jennifer McWhirter -- examine photographs and other images as artifacts created by research subjects that both researchers and subjects can regard and interpret.

In editing this volume, Prosser has created an excellent introduction to image-based research, but the book also reflects some of the changes I noted earlier.  Photography is the dominant, but by no means only form of visual imagery examined.  Traditional social science disciplines frame some chapters, but not others, and a variety of other intellectual communities appear and disappear as the reader moves through the book.  Claims made for the realist value of image-based work stand close by critiques of ascribing either factuality of unambiguous meaning to particular images, and the authors of these essays have followed a wide variety of career paths.  Taken together, Prosser's book reveals that intellectual projects of image-based research are well underway in many quarters, that they remain both promising and problematic, and that there's no end in sight. 

A reader looking for a single, coherent, all-consuming statement of the field would be hard pressed to find it in Image-based Research.  Within its diverse chapters, there is much to engage with, but we’re left with the impression that the "field" of image-based research has a pretty ragged edge, an uncertain center, and a bit more complexity and ambiguity than many disciplinary scholars might like.  Perhaps a more coherent statement of visual research would help.  And yet. . . 

"Researching the Visual"

Researching the Visual, by Emmison and Smith appeared two years after the volume edited by Prosser.  The authors do not refer directly to Prosser's book, nor is it listed among the references they cite.  However, in some respects Emmison and Smith set out to challenge studies of the sort accounted for by Image-based Research; in other respects they also overlook them. 

The authors of Researching the Visual are both faculty members in a department of anthropology and sociology at the University of Queensland (Prosser, by the way, is on the Education Faculty at the University of Leeds).  An obvious contrast with Image-based Research is that Researching the Visual is not a diverse collection but a monograph authored wholly by Emmison and Smith.  Their consistent voice and attention to a central, unifying argument gives the book a quite different feel.  This also creates different opportunities for analysis and a different problematic for readers and reviewers. 

Emmison and Smith begin their treatise with a thoughtful review of prior efforts to articulate the role of visual imagery and the visual in the work of sociologists, anthropologists, semoticians and others "interested in social and cultural processes."  In arguing for the centrality of the visible in analyses of this sort they distinguish between studies of "the visible" and studies of particular "images," championing the former as analytically more productive and defensible than the latter.  In their own words,

What needs to be considered. . . is the way in which the visible features of the social world, which are readily available to the naked eye -- not their representation in photographic images -- constitute data for investigation.  Stated in its simplest form, we are proposing that visual data should be thought of not in terms of what the camera can record but of what the eye can see (p. 4).

 This is a good argument, one that John Grady made a few years earlier in his own, somewhat parallel conception of "visual sociology" (1996).  As promoted by Emmison and Smith, the substantive implications of this argument "position visual research as a central theme of investigations into society and culture.  It allows us to demonstrate that scholars in a diverse range of areas (not just those looking at photographs) have been 'doing the visual'" (p. x).  Among the early scholars they refer to in this regard Georg Simmel figures prominently, as do Chicago human ecologists who conducted a diverse array of visually and spatially oriented studies during the first half of the 20th Century.  More recent references are to the visually sensitive writings of Erving Goffman and to studies by Lyn Lofland (1973) and Edward Hall (1966).  

In celebrating this broad conception of visual research, Emmison and Smith note its potential to revitalize an equally broad notion of social research. 

Ideas about display, status and interaction allow us to tap into the rich vein of Goffmanian interactionism.  Ideas about surveillance, visibility and privacy bring to mind Elias and Foucault.  Readings of objects, buildings and places allow reference to Levi-Strauss on nature/culture or to postmodern theory on architecture and so on  (p. ix).

As a corollary to this celebration, the authors of Researching the Visual criticize "self-proclaimed visual researchers" for focusing too much on images, still photographs in particular, an orientation that truncates the visible dimensions of social life.  They add to this criticism the contention that those sociologists, anthropologists, and ethnographers who have used photographs have done so in "a purely illustrative or documentary fashion" that falls far short of the kind of theoretical and analytical inquiry required to advance knowledge in the field.  While the nature of this "field" is not always clear, Emmison and Smith present their own orientation as a corrective to these deficiencies and challenge their readers to move "beyond the image in general, and the photograph in particular. . [so that] visual research can become a powerful and theoretically driven domain of social and cultural inquiry" (p. 55).

This is a provocative line of reasoning, one worth attending to not only by "self-proclaimed visual researchers" but by sociologists, anthropologists, and others who have heretofore regarded their disciplines in essentially non-visual terms.  Scholarship is indeed diminished when we overlook the power of “seeing” in trying to understand culture and social life, or when we isolate questions about the visible dimensions of social life from questions about economies, language, socialization, kinship systems, and so on.  And it is not that difficult to find instances in which the critique Emmison and Smith make of still photography is right on target. 

Having said that, the rhetoric with which Emmison and Smith advance their arguments is at times so polemical than it either detracts from the points they want to make or displaces deeper analysis.  In one of several strongly worded statements about their intellectual purpose, for example, Emmison and Smith write, "The central theme of this book is the need for visual research to break free of the tyranny of the photograph" (p. 107).  Some social scientists may indeed wax romantic about the value of still photographs, but does that constitute a "tyranny"?  Similarly, while Emmison and Smith note from time to time that they are not trying to displace the appropriate use of photographs in social research, they also take every opportunity to insist that visual research can be done quite well without it.  In most cases they are technically correct, but many studies they cite in this regard could be pursued more economically, systematically, or effectively by using photography and other image recording technologies.  Why not say so?

Emmison and Smith do suggest that some image-based and photograph-assisted research strategies make sense, primarily those in which visual media have been used as a relatively mundane technology for recording the visible world.  And they comment with enthusiasm about the work of Bateson and Mead (1942)  and about Goffman's 1979  book Gender Advertisements (though they read the word "advertisements" literally).  But Researching the Visual seems much less oriented towards encouraging more of this work than it does to challenging the value of image-based approaches -- Doug Harper (1993)  is cited as a leading proponent -- in which photographs are intended by researchers to be aesthetically engaging or illuminating, as well as informative.  The broad range of image-based approaches found in Prosser's book makes these priorities seem misplaced.  And, while Emmison and Smith try to distinguish between the approaches they regard as appropriate (i.e., prosaic and essential) and inappropriate (i.e. poetic or convenient), they sometimes confound the two, and when they do, seem almost eager to throw out one with the other. 

This eagerness appears to rest in part on a somewhat unrealistic methodological stance, one that lacks sensitivity to the interplay of aesthetics, convenience, and logic in the design and conduct of social research.  Emmison and Smith are quite right to champion logic, but wrong to suggest that research of any sort can be done without making aesthetic choices.  Consistent with this distortion, they want to radically circumscribe the representational value of photographs to social inquiry.  "Unlike other forms of storing information,” they write, “photographs are signs which bear an iconic resemblance to the reality they describe. . . Yet this should not negate their fundamental similarities with completed surveys, notebooks and so on as storage devices." 

In restricting their representational virtues for social research, Emmison and Smith reject photographs in terms that are just as romantic as those of the researcher-photographers they criticize for over using them.  Perhaps they have the wrong analogy.  A more apt comparison than with the survey form or notebook appears between the photograph and the audio tape.  Just as linguists or ethnographers might use a sample of transcribed, audio-tape-recorded speech to illustrate and exemplify an important concept, or to compare speech patterns over time or across settings or speakers, or to systematically look and listen many times over to a segment of conversation, so too a visual researcher might use photographs or video recordings.  Researcher observations and note-taking also can be extremely valuable to this kind of work.  However, audio and visual recordings have some distinctive features -- i.e., their capacity to simultaneously illustrate and exemplify and the opportunities they generate for systematic comparison and for repeated analysis of an incident or setting -- that make it possible to examine research questions that would be very hard to investigate without them.  But that's not all.  Just as some audio recordings allow us to hear people making sounds (narratives, dialogues, songs, stories, etc.) in ways that have an aesthetically affecting and enduring impact, so too some visual recordings allow us to see people doing things in similarly enriched terms.

The force with which Emmison and Smith reject the latter kind of representation distracts them from a more prosaic and compelling explanation for why photographs, drawings and other two-dimensional representations have fit better with visual social research than some of the "visible" settings, bodies, artifacts and other objects of inquiry they regard as more essential.  Doesn't the dominant role of printed paper in distributing research reports have something to do with this?  Indeed, while Emmison and Smith extol the virtues of researching visible objects of various sorts, what we find in Researching the Visual, the book, is limited to text, a few photographs and a few line drawings.  The "things" themselves they want us to attend to more directly don't fit because they've chosen to write a book -- and not make a box, construct a room, or parade a body.  They describe some of these "things" in written text, and Emmison and Smith are quite deliberate in accounting for how and why they do so.  But don't "texts" have some of the same potential to distract us from "seeing" that the authors attribute to photographs and other images?

A parallel inclination to over-dramatize and under-analyze appears when Emmison and Smith refer to issues of "visual depiction or representation."  They place these issues outside the ken of most social scientists, particularly those working in North America, and specify further that "studies of visualization in science have been largely confined to the physical and life sciences."  As I noted earlier, however, during the last twenty years scholarship about representational strategies, tools and conventions -- including the imagery used by social scientists – has flourished among linguists, literary critics, anthropologists, and  sociologists, not only in Europe but in America as well (Becker, 1986; Gusfield 1981; Hunter 1990). 

In neglecting this scholarship, Emmison and Smith's account of Howard Becker’s work has ironic overtones.  They refer to Becker only as the person who got visual research off on the wrong foot some 30 years ago by noting a correspondence between sociology and still photography, and they seem completely unaware of his continuing scholarship about other forms of visual representation and the implicit imagery of social theory -- key themes of their own analysis.  As Becker notes in  "Imagery," the lead chapter of his most recent book (1998), "[Herbert] Blumer thought, and so do I, that the basic operation in studying a society -- we start and end with them -- is the production and refinement of an image of the thing we are studying" (p. 12).  While statements of this sort clearly support some of Emmison and Smith's central arguments, they also suggest that the authors of Researching the Visual have more intellectual company in advancing the case than their rhetoric implies.

Once they get their somewhat overstated arguments on the table, Emmison and Smith thoughtfully explicate strategies for studying four different kinds of "visual data," each  re-stated as a title for one of the last four chapters in their book.  In Chapter Three, "Two-dimensional Visual Data," they review the prospects for conducting both quantitative and qualitative studies of advertisements, cartoons and comic strips, headlines, signs and maps.  In the chapter that follows, "Three-dimensional Visual Data," they examine visual studies of people, settings and objects -- including cars, houses, gravestones and rubbish.  Chapter Five focuses on what the authors call, "Lived Visual Data," and gives special attention to the built environment and to the movement of people in time and space.  In "Living Forms of Visual Data," the title of their concluding chapter, Emmison and Smith describe studies in which the human body and its accessories provide visual indications of social activity and social position.

Within each of these chapters, Emmison and Smith make useful distinctions in how different configurations of the "visible" can be noticed, accessed, and examined, and they refer to numerous studies conducted by other researchers to illustrate each configuration.  They also provide a set of exercises  for students and other prospective researchers (these appear in shaded side-bars that accompany the main text for each chapter).  Many exercises are intriguing.  Some would make great projects for students to undertake in an undergraduate class, or even an introductory research methods class for graduate students, and some could be extended into full-scale thesis or dissertation studies.

The combination of argument and critique about the status and scope of visual research and exercise sets for students is at times helpful, at other times confounding.  But the range of exercises -- and the rationale for limiting this range – does not always square with Emmison and Smith’s attention to student readers.  The neglect of camera work in field studies seems particularly odd, as does the authors’ contention that studies involving video tape and mass media are too complex and challenging for students.  Emmison and Smith are quite open about their intent to demonstrate that even without camera work, video work and mass media studies, students can be engaged in interesting investigations of "the visual."  They are right, of course, but it's not clear who they are arguing against.  And they push so hard away from photographs, video tape and the mass media as to handicap students who might come to their book looking for a full complement of visual research tools. 

Something similar occurs in how Emmison and Smith approach covert and overt research with human subjects.  The preference they show for covert research is not so much justified as it is consistently affirmed in phrases such as this: "The great bonus of these very visible sources of information is that they allow us to explore social life covertly" (p. 110).  However, their guidance on this point is contradictory.  They note that, "we have to be cautious in assuming that images have fixed meanings or ideological effects just because we 'experts' are able to read these into them,"  and that "an ability to correctly read some kinds of images (such as religious art) can take years of socialization" (p. 67).  These cautions are sound advice, but they seem ingenuous when juxtaposed with the authors’ contention that, "Whilst ethnography or interview research is sometimes required to flesh out the meanings of objects and improve our inferences, we can often get by without it" (p. 110).  My own research supports a quite contrary conclusion, and Emmison and Smith provide little if any guidance about when this kind of inquiry is essential and when it is not.  Nor do they suggest how to proceed with ethnographic study when we can't "get by" without it. 

Is it their preference for covert research that leads Emmison and Smith to neglect the representational value in research reporting not only of images, but also of native texts?  They seem unconvinced that images and texts drawn from raw data files can enrich a reader's understanding of the people and places implicated in a research study.  Or that as rhetorical devices, these elements can also enliven a research report and provide an account that is more "multi-vocal" or that offers alternative "points of view."  Nor do they suggest that by using "data chunks" in their reports, researchers provide opportunities for readers to reassess their claims.  And yet, despite their lack of enthusiasm for these prospects,  Researching the Visual does exemplify both the value and problematic of presenting just such materials. In commenting on a pair of "destabilized" advertisements created by one of their students, for example, the authors' explain that "the text is constant and it acquires its (different) meanings through the changing images with which it is juxtaposed."  However, because an image of this student's work appears in their book, I could examine it myself.  And when I did, I came to the quite different conclusion that the meaning changed at least in part because new text was added and old text deleted. 

Part of the problem with these and other loose ends is that Emmison and Smith frequently paint with quite a broad brush.  As yet another example, they contrast "European traditions of visual inquiry where photographs are images for interpretation and decoding" with the work of American social scientists for whom, "photographs are invariably images which are taken by the researchers" (p. 24).  But this contrast has less to do with Americans and Europeans than with field researchers and investigators engaged in other kinds of studies.  If we compare Barthes or Bourdieu with Sontag  -- or with Richard Chalfen (1987), Michael Lesy (1980), or Alan Trachtenberg (1989) -- the American and European contrast they've made falls apart, just as it does if we compare researchers on either side of the Atlantic who use photographs or video recordings as tools for conducting field research. 

Giving Emmison and Smith the benefit of the doubt, their overstatements and broad brush strokes appear as a kind of provocative enthusiasm for the ideas they are bringing forward in this book.  But their analysis and rhetorical stance also represent an interesting manifestation of the structural issues I noted earlier: the growth of methodological communities, the tensions between disciplinary and interdisciplinary study, the movement of sociologists-by-training through applied and allied disciplines, and the development of new academic programs and lines of inquiry, some of which may be only loosely related to traditional social science disciplines and within which no one is quite sure where visual research can best be placed.  While Emmison and Smith chart a provocative path through this complex, dynamic enterprise, they are also very much a part of what they write about.

Tracing the Visual in Social Research

Image-based Research and Researching the Visual  both identify and illustrate a range of visual approaches that sociologists and other social scientists might want to consider in enriching their investigations of culture and social life.  Both provide examples of studies that could be extended productively to other settings, other populations, and other objects of inquiry, and both argue for giving greater attention to the visual dimensions of culture and social life. 

Image-based Research presents this case in a way that is generous towards the use of images, encouraging, thoughtful, and eclectic.  Readers will be rewarded by encountering several kinds of theorizing in which imagery and the visual are front and center -- with photographic media getting special, but not exclusive, attention -- and a range of pragmatic applications.  In Researching the Visual, Emmison and Smith argue for broadening visual research beyond the kind of image-centered and photographic work represented in Image-based Research, and their arguments and examples make a good case for doing just that. 

In some respects the two books are complementary.  For example, some contributors to Prosser book express concerns about social researchers moving too far away from their research subjects; Emmison and Smith seem far more worried about those who might get too close.  Beyond their particulars, however, the central thrust of these two books define something of a paradox:  Image-based Research suggests that attending to images can expand our social scientific field of view, but Researching the Visual suggest just the opposite, that images can narrow what we look at and are inclined to see. 

Both, of course, are right,  This becomes clear when we examine four different romances embedded in these two texts, each of which can distract sociologists and other social researchers from sound strategies for studying the visual dimensions of culture and social life: the romance of the image: the romance of the discipline; the romantic rejection of the discipline; and the romantic rejection of the image.  Let me comment on each of these in turn.

As a corrective to those who can find everything they need in images, both of these books (but Researching the Visual in particular) reveal visual data to be much more varied, complex and ubiquitous than photographs or other fixed "images" per se.  Indeed, the range of visible phenomena and visual indicators examined in these two books is considerable.  This range positions visual referents as a central, not peripheral, feature of culture and social life.  It also recommends greater visual sensitivity in investigating phenomena of continuing interest to sociologists and anthropologists.  These include families, beliefs and attitudes, organizations, forms of work and association, and education -- not just art, clothing, housing and other things that might appear "intrinsically visual."  Photographs themselves can certainly play a role in encouraging this kind of sensitivity, but so can other kinds of images, artifacts and media.  And, as Emmison and Smith argue, direct observation can play a role as well. 

That said, neither of these two books examines how far beyond photographs, line drawings and other fixed images social researchers are willing to move in challenging the reporting conventions of their own scholarly communities.  When books and paper predominate for the latter, it is unlikely that video tape, objects, installations, and visual performance pieces can contribute much more than they do at present to research reporting.  That no doubt limits the kinds of data social researchers are most comfortable working with -- though, as Emmison and Smith argue, even research reports that take the form of un-illustrated, written texts can be based on close analysis of varied visual data.

The books by Prosser and by Emmison and Smith also reveal how the intellectual thrust of visually-sensitive social research breaches the boundaries of core social science disciplines.  Disciplinary romantics and purists may find this troubling.  However, it's worth noting that neither sociology nor anthropology (nor cultural studies, for that matter) is inert or unchanging.  The disciplines have endured not as fixed and frozen objects but as dynamic communities that extend across time and space through a variety of representation media.  None has a monopoly on scholarship, on knowing about the world, or on models for understanding culture and social life.  As Dell Hymes (1974) put it, scholars must " reinvent" their disciplines on an ongoing basis as new ideas, perspectives, and objects of inquiry -- including images and the visible world -- move into or out of focus.  Disciplinary scholars who eschew this kind of re-invention are at a disadvantage in developing the kind of visual sensitivity recommended by Prosser and by Emmison and Smith.

Characterizing the disciplines as dynamic does not mean they are formless or irrelevant to visual research.  If by discipline we mean a distinctive set of intellectual purposes and the crafts necessary to pursue them in ways that are well-informed and thoughtful, then sociology, anthropology and the other social sciences are quite alive and well -- not only as professional communities but as resources for the conduct of image-based or visual research.  And the purposes of such disciplined inquiry are well worth fighting for in deliberate, "disciplined" terms.

This suggests that anti-discipline romantics are no better off than disciplinary romantics in making sense of either visual social research or of culture and social life.  For example, one consequence of the well-documented, polysemic character of objects and images is that we can't determine what these things invariably mean to "people" in general.  Rather, we need to investigate their meaning for different people in different social contexts.  This leads us to approach with great suspicion analyses by social researchers and others who assign only their own meanings to what they see.  The discipline of informed, systematic social research generates these consequences and suspicions.  It also leads us to recommend ways of investigating images, imagery, and visible social life that are systematic, situated, comparative, and empirical.  Sociology has much to contribute in developing disciplined strategies of this sort, and much to gain in return.

This does not necessarily mean that establishing "the image" as a stand alone specialty of sociological or anthropological investigation makes good sense.  What would we include?  Photographs, movies and video tapes seem likely candidates and so does the full range of "visible" phenomena examined by Emmison and Smith or by Prosser.  But what about FAXes, scans and machine-made photocopies?  These too are images, products of a form of "camera work" used routinely in both social life and social research.  The pages of this journal are also images, accessible only to those who can see and read this particular typography -- English words, not too small, too faint or too blurred, etc.  All this is well worth investigating, but maybe not all in the same way.  As Emmison and Smith note about the extraordinary range of subjects that can be depicted in photographs,

. . the apparent unity these phenomena have as a consequence of being thus collected [i.e. photographed] is entirely spurious.  Their utility as data does not stem from their character as photographs, but lies elsewhere.  To appreciate what their utility might be we have to focus on the characteristics of these phenomena as objects in their own right" (p. 3).

A better bet is to recommend visual recording media more routinely as tools for conducting social research -- as routinely, for example, as scholars currently recommend audio or data tape recorders, survey questionnaires, notebook computers, and yellow pads of college-ruled paper.  However, this recommendation frequently runs up against the fourth unproductive romance I noted earlier, the rejection of images as evidentiary and representational resources. 

Are images themselves somehow too messy and fugitive to support social inquiry?  If not, are they nevertheless too loaded with symbolism and promise to use in research reports?  In trying to answer these questions, it is important to acknowledge that images are no more fixed and tidy than products of the other recording tools used by social researchers.  To use them reliably, people need to learn how to work with them -- how to use them appropriately and effectively in collecting and analyzing data and in teaching others what they've learned through this kind of work.  This is not a simple task, but neither is it a simple task to write field notes, construct a set of interview questions, "clean up" census track data, or explicate key concepts in a social theory text.  Using any one of these research tools requires a disciplined approach to inquiry, one in which intellectual purpose and technical means are well matched.  Romantic acceptance or rejection of the image -- or of an academic discipline -- makes that more difficult.

The four unproductive romances I've distilled from these two books were not unknown twenty years ago when Images of Information came to press, but neither were they configured in quite the same way as they are now.  Romantic views of disciplines were stronger then; romantic views of the image may be stronger now.  And a close look at the various threads that feed into and out of these romances does not reveal a straight line of continuing advances or decline, but rather a spiral of inquiries in which texts, images and visual sensitivity have taken a few more turns.  Sometimes this spiral has taken scholars out of sociology or the other social science disciplines and led them into new intellectual venues, sometimes it has brought them back in.

If this characterization is accurate, the central axis of the spiral I'm referring to -- the evolving center of intellectual work that has brought us Image-Based Research, Researching the Visual, and a rapidly growing list of related books and new journals -- cannot be plotted entirely from within sociology or anthropology itself, nor, for that matter, from within the qualitative research or cultural studies communities.  Rather, the investigations that define this spiral fall only in part within each of these domains.  And when work contributing to this larger intellectual project does appear to scholars within a disciplinary community, it may do so as a relatively short and isolated line of inquiry.  Each of these lines may reflect considerable energy and intellectual promise.  But they may also fade out of sight rather quickly within a given discipline,  as momentum carries both scholars and their work into neighboring domains.

Ignoring the cross-disciplinary and dynamic axis of visual social research presents liabilities both for new disciplines and for old.  If new intellectual communities affirm the visual as a rejection of traditional disciplines, they lose the opportunity to build on the considerable contributions those disciplines have made to visual inquiry -- even if such contributions are not central to the traditional discipline itself.  Similarly, if sociology and other traditional social sciences ignore the kind of visual inquiries examined in Image-based Research and Researching the Visual, these disciplines will appear less and less useful as scholarly orientations for doing this kind of work.  The latter is something many sociologists could live with.  But the corollary may be less agreeable, that the romances noted above may lead sociologists to overlook visual tools and strategies necessary to more fully apprehend and theorize about the social world.  That's both an outcome and a trajectory that sociologists of any stripe ought to want to avoid.

References

Bateson, G. and Mead, M.  (1942).  Balinese character: A photographic analysis.  New York: New York Academy of Sciences.

Becker, H. S. (1974). Photography and sociology.  Studies in the Anthropology of Visual Communication, 1: 3-26. 

Becker, H. S. (1986). Writing for social scientists. Chicago, IL: University of Chicago Press.

Becker, H. S. (1998). Tricks of the trade:  How to think about your research while you are doing it. Chicago: University of Chicago.

Chalfen, R.  (1987).  Snapshot versions of life.  Bowling Green, OH: Popular Press.

Clifford, J., & Marcus, G. E. (1986). Writing culture: The poetics and politics of ethnography. Berkeley, CA: University of California.

Collier, J. J. (1967). Visual anthropology: Photography as a research method. New York: Holt, Rinehart & Winston.

Emmison, M. and Smith, P.  (2000).  Researching the Visual.  Thousand Oaks, CA:  SAGE

Erikson, K. (1976). Everything in its path. New York: Simon and Schuster.

Filstead, W. J. (1970). Qualitative methodology: First hand involvement with the social world. Chicago: Rand McNally.

Fleischhauer, C., & Brannan, B. W. (1988). Documenting America, 1935-1943. Berkeley, CA: University of California.

Goffman, E. (1979).  Gender advertisements.  New York: Harper and Row.

Grady, J. (1996). The scope of visual sociology. Visual Sociology, 11, 10-24.

Gusfield, J. (1981). The culture of public problems: Drinking-driving and the symbolic order. Chicago: University of Chicago Press.

Hall, E. T.  (1966).  The hidden dimension.  New York: Doubleday.

Harper, D.  (1993).  On the authority of the image: Visual sociology at the crossroads.  In Denzin, N and Lincoln, Y.  Handbook of qualitative research.  Beverly Hills, CA:  Sage, pp. 403-12.

Hunter, A. J. (1990). The rhetoric of social research: Understood and believed. New Brunswick, NJ: Rutgers.

Hymes, D. (1974). Reinventing anthropology. New York: Vintage Books.

Jackson, B. (1977). Killing Time: Life in the Arkansas Penitentiary. Ithaca, NY: Cornell University.

Kerr, C. (1962). The uses of the university. New York: Harper Torchbooks.

Lesy, M.  (1980).  Time frames: The meaning of family pictures.  New York: Pantheon.

Lofland, L. (1973).  A world of strangers.  New York: Basic Books.

Mehan, H. (1979). Learning lessons: The social organization of classroom instruction. Cambridge, MA: Harvard University Press.

Prosser, Jon. Ed. (1998).   Image-based research: A sourcebook for qualitative researchers.  Bristol, PA: Falmer.

Stasz, C. (1979). The early history of visual sociology. In J. Wagner (Ed.), Images of information: Still photography in the social sciences (pp. 119-136). Beverly Hills, CA: SAGE.

Trachtenberg, A.  (1989). Reading American photographs: Images as history, Mathew Brady7 to Walker Evans.  New York: Hill and Wang.

Van Maanen, J. (1992). Tales of the field. Chicago, IL: University of Chicago.

Visual sociology: Expanding sociological vision, The American Sociologist, [___]

Wagner, J. (1979). Images of information: Still photography in the social sciences. Beverly Hills, CA: SAGE.

Whyte, W. H. (1972). "Please, just a nice place to sit.", New York Times Magazine.