Does image-based field work have more to gain from extending or from rejecting scientific realism?

A review of Robert Coles, Doing Documentary Work and
Sarah Pink, Doing Visual Ethnography

Jon Wagner
University of California, Davis

Published in   Visual Sociology 16 (2), pp 7-21, 2002

The social psychologist William James reputedly held that there were two different but equally valid ways to go about "doing" science.  You could try to avoid error and then seek the truth, or you could first seek the truth and then try to avoid error.  James' characterization of a century or so ago is a good reminder of two first principles:  All science involves a search for relatively reliable ways of identifying, observing, examining and talking about things that at least appear to be objective, external realities;  and all science also depends on the subjective experience and knowledge constructing talents of individual researchers and scientific communities.

Ideas about image-based field work can easily get caught up in polarized advocacy of one or the other of these two principles. But photographs, drawings, paintings, video tapes and other cultural artifacts -- houses, clothing, books, computer screens, lab equipment, paper print-outs of survey data or experimental trials -- belong to both.  This generates two abiding ambiguities:  The first ambiguity refers to how an individual image or artifact can and should be read -- as an explicit, precise, and matter-of-fact communication or as a polysemic and ambiguous social and cultural artifact.  The second ambiguity refers to how images in general can and should be used in social inquiry -- as information-rich data for extending scientific investigations or as evocative artifacts for challenging or stepping away from a science too narrowly conceive. 

Robert Coles and Sarah Pink have much to say about these ambiguities and their implications for image-based field work.  As authors of Doing Documentary Work (Coles, 1997) and Doing Visual Ethnography (Pink, 2001), they both see in the evocative and subjective power of images a potential to enrich social inquiry and representation.  Both also reject the "realist romance" of conventional social science and favor other approaches to image-supported inquiry.  But the power and promise that Coles and Pink find in image-based work also stems in part from considering images in "realist" terms.  This generates a continuing tension between how these two authors theorize about documentary work and visual ethnography and the insightful observations they provide of "doing" that kind of work.

Coles and Pink have both worked with text and images in inquiries of their own, and they draw upon this work to provide stimulating accounts of using film, photography, video, and writing to study and represent social realities.  Among the more important issues they examine are those shaped by theoretical and practical challenges of negotiating different forms of objectivity and subjectivity, fact and fiction, science and art, image and text, and realism and reflexivity.  Dualities of this sort have received continuing attention within the social sciences, but the candor and craft knowledge that Coles and Pink bring to bear offer fresh insights into how such issues can be conceptualized and managed to support productive field work. 

This is not to say that the two books cover the same ground.  Parallels are evident in how Coles and Pink approach the particulars of image-based field work, but so too are contrasts.  Indeed, each author takes for granted key aspects of this work that the other regards as problematic.  This makes Doing Documentary Work and Doing Visual Ethnography an instructive pair to examine together.  What do the two kinds of "doing" that Coles and Pink write about entail, require or produce?  Within what traditions of inquiry are they best positioned and assessed?  And to what extent do they require a willingness to extend, suspend or reject principles of conventional scientific practice?

Doing Documentary Work

Doing Documentary Work is organized around an introduction and four individual essays first presented by Coles as a lecture series sponsored by the New York Public Library and Oxford University Press.  As an epilogue, the book also includes Coles' proposal for an academic course on "documentary studies." 

Through the introduction and first essay, Coles offers a brief but thoughtful overview of what he calls "documentary work."  The overview leads to a second essay focusing on moral and psychological tensions and to a third in which Coles examines the interpenetration of "fact and fiction."  In the final essay, Coles steps back to examine documentary inquiry as a broadly defined but distinctive form of field work communicated through varied literary, filmic, and photographic genres. 

Coles starts with James Agee and Walker Evans' well-known documentary study, Let Us Now Praise Famous Men (1960 [1939]) and uses Agee's own prose to frame key challenges of documentary work.  Ethical issues are always close at hand in Coles' writing, but Coles also places front and center the complementary value of empirical and interpretive modes of inquiry.  In affirming the former he notes,

The word documentary certainly suggests an interest in what is actual, what exists, rather than what one brings personally, if not irrationally, to the table of present-day actuality.  Documentary evidence substantiates what is otherwise an assertion or a hypothesis or a claim.  A documentary film attempts to portray a particular kind of life realistically; a documentary report offers authentication of what is otherwise speculation. (p. 5)

But Coles is quick to add the complicating observation that "a search for the factual, the palpable, the real, a determined effort to observe and authenticate, and, afterwards, to report, has to contend, often enough, with a range of seemingly irrelevant or distracting emotions -- the search for objectivity waylaid by a stubborn subjectivity" (p. 5).

To many modern and post-modern scholars, the divide Coles locates between  objective and subjective realities represents a challenge to the epistemological foundations of science itself, and extremist views of this sort appear routinely in two widely supported  counter claims:  From advocates of modern, or so-called "traditional" scholarship we hear that subjectivities must be stamped out to protect the empirical integrity of scientific inquiry.1  Avoid error first, they say, and only then seek the truth.  From postmodern critics we hear that the inevitable subjectivity of human activity -- science included -- not only negates the possibility of wholly objective inquiry but defines both the foundation and ceiling for whatever knowledge one can reasonably construct.  Seek subjective truths first, they argue, because there are no alternatives.

Coles side-steps these contrary romances by approaching this divide pragmatically.  As a practiced field worker he appreciates the value of documents that are richly detailed, informative and evocative.  Instead of an epistemological dilemma, he refers to subjective/objective polarities as a pragmatic challenge distinctive to the craft of "doing documentary work."  As Coles puts it, "This mix of the objective and the subjective is a constant presence and, for many of us, a constant challenge -- what blend of the two is proper, and at what point shall we begin to cry 'foul'"? (p. 8). 

In the balance of Doing Documentary Work, Coles examines how this "mix" of the subjective and the objective shapes complementary tasks of researching and representing social, interpersonal and cultural realities.   He looks closely at the craft with which Dorothea Lange selected and cropped her own photographs.  He reprises Walker Evans' commentary about the subjective and objective dimensions of his own craft, noting the subtle distinction Evans made between "documentary" work and art produced in a deliberate "documentary style."  In reviewing the documentary films of Frederick Wiseman, Coles describes how Wiseman applies the varied crafts of film maker, editor, visual poet, social critic and ironist in "probing social reality, yes, but also arranging it, composing it, as artists or writers do" -- in Wiseman's case, to place viewers of his films "in the midst of a particular social and institutional scene."

Among the writers Coles examines through his documentary lens are William Carlos Williams, James Agee, George Orwell, Oscar Lewis, Studs Terkel, John Baskin, Anthony Wallace and Kathleen Norris.  Among the photographers and film makers are Wiseman, Evans, Lange, Robert Frank, and Thomas Roma.  But Coles seems to reserve his deepest appreciation for documentary works that bring together text and images, as Wright Morris did on his own or as Agee and Evans, and Dorothea Lange and her husband, the economist Paul Schuster Taylor, did in partnership with each other.

Coles' attention to how these "documentarians" have struck a workable balance between subjective and objective realities is an insightful brief for enriched work with visual materials, and Coles' observations are all the more valuable for defining the latter in broad, common sense terms.  As part of his own field investigations, for example, Coles has interviewed children by working with them as a kind of adult drawing partner, communicating through and about the pictures they both make with crayons and paper (Coles, 1992).  Coles also writes vividly of what he "sees" and what others tell him they have looked at, placing visual matters close to the interpersonal and phenomenal world of those he writes about, and he has both supported and collaborated with photographers (Coles & Nixon, 1998).

This attention to the "visual" applies not only to the children and adults Coles has studied, but to those he has studied with, including William Carlos Williams and Erik Erikson.  In Doing Documentary Work, accounts of his conversations with Erikson fade in and out of Coles' observations about other writers, photographers and film-makers.  In one of these, Erikson describes how the visual substance of psychoanalytic craft has increased since the time of his own training with Anna Freud. 

Back then in our psychoanalytic conferences, everything was talk: the patient said this, and that, and we had this or that to say in response. . .  As I look back at that time and compare it with our time, I realize how different this work is. Now our eyes are constantly being given something to do!  Back then, many psychoanalysts talked of closing their eyes, while they listened, listened very intently.  Now, we look and look and look -- and our patients are telling us that they saw this, they saw that, at the movies and on TV and in the magazine advertisements. (p. 231)

Coles characterizes Erikson not only as a mentor with sophisticated visual sensibilities, but as a documentarian in his own right.  He recounts another conversation that occurred when Erikson was working on his biography of Ghandi.  At the end of a thoughtful exchange about the evidential and fictive attributes of "stories," Erikson exclaims, "Now, you see why I want to go to India and interview those people who knew Ghandi and worked with him!  You see why I want you to show me how you use your tape recorder!"  (p. 142)

While Coles clearly celebrates authors and creators of visual and textual works he has chosen to review, his larger purpose is to examine the distinctive challenges of "doing documentary work."  This is an intellectually useful task, and Coles is an ideal candidate to take it on.  He reads widely, writes with sensibility and clarity, and faces up without apology to the moral, interpersonal and political complexity found in field work of any sort.  Coles also paid his own dues as a field worker and learned much in the process -- documenting with his tape recorder and note book the late night Mississippi meetings of political activists in the 1960's or the lives of children in a wide range of social and cultural settings.  Coles has reported on his investigations as the author of several dozen books well known to both scholars and lay readers, including the "Children of Crisis" series for which he won the Pulitzer Prize.  He has considered carefully the ethical issues of representing in public the details of private lives, and he's been active as a teacher, lecturer, and editor with a special interest in children and in documentary work itself.2

Having said that, Coles' enthusiasm for documentary work overshoots a bit his analysis of what it takes to do it well.  The mis-match appears not so much from neglecting complexities of the work he values, but rather from Coles' reluctance to distinguish interpretation more cleanly from empirical observation, and a corresponding eagerness to separate documentary work from an overly narrow notion of scientific practice.  Taken together, these two dispositions truncate Coles' treatment of documentary craft.  He writes far too little about how the evidentiary value of documentary work can be assessed or encouraged, and leaves the door open for readers to define "good" documentary work as nothing more or less than documentary "works" of great renown. 

This liability is attenuated by the pains Coles takes to keep complexities of field work in the foreground of his discursive and probing prose.  And yet, while Coles' thoughtful questions represent subjective inquiry as a normal feature of documentary practice, they are less likely to normalize the "objective" dimension that Coles says he also values. 

This asymmetry is unfortunately echoed in Coles' comments about social science itself -- none of which are sympathetic -- and it is augmented when Coles gives voice to other critics who share his views.  As one example of the latter, Coles includes a verbatim comment from the civil rights leader Bob Moses about Coles' own "scientific persona" as it appeared to Moses during the Mississippi voter registration project of the mid 1960's:

Here, so many of you folks are trying to explain everything, everybody with these 'general laws,' -- you impose ideas on people.  That's what a professor of mine said is wrong about a lot of social science -- and the newspaper and magazine editors who take it so seriously: the 'rage,' he said,' to reduce,' to simplify, to explain with a definition or conclusion that is supposed to include, to take care of, to account for everything.  (p. 37)

A related critique appears in Coles' account of his own socialization into and out of scholarly publishing, in this case with Coles as critic, rather than the exemplar of misguided science.  When an editor tells Coles that the word "poignant" is too subjective and should be removed from a manuscript he'd submitted for publication, Coles bridles: 

I remember telling my editor friend that all of the 'research' I had written up for this 'paper' was 'subjective' -- an estimate or interpretation on my part of what I thought I had seen and heard happening in the lives of children, in their minds, rather than a chronicle of what happened independently of my mind (an account of the unfolding of an objective series of events). (p. 28)

The editor's response, understandably enough, was to say, "Yes, but there are degrees."  But Coles neither presents that response as useful wisdom, nor does he explore further his own understanding of what "degrees" of objectivity might entail.  This is unfortunate, for Coles has made creative and systematic use of tape recordings, artifacts, direct observation and field notes throughout several decades of his own labor-intensive field research.  His writing may be evocative and personal, but it is also well-informed by detailed, empirical observation.  From this work and from his comments about other documentarians, I'm convinced that Coles, too, judges some studies or reports as more or less accurate or authentic than others.  But he doesn't have much to say here about how to exercise judgment of that sort. 

This shortcoming is not a fatal flaw, and, hopefully, Coles will someday bring to authentication issues the same quality of thought he has brought to other dimensions of documentary work.  Among the latter, Coles' examination of ethics is exceptionally rich for questions he raises about field work in general.  "What are one's obligations not to oneself, one's career, the academic world, or the world of readers," he asks, "but to the people who are, after all, slowly becoming not only one's 'sources' or 'contacts' or 'informants,' but one's graciously tolerant and open-handed teachers and friends. . ." (p. 61-62).  Somewhat later he states the query in more pragmatic terms:

More bluntly, what, if anything do we owe those we have 'studied,' whose lives we have gone to document?  Should we, for instance, send back the writing, the photography, the film once it is completed?  If so, at what stage of that work's development: as it is being assembled, as it is being edited, before it is published or exhibited or shown on television or in a movie house, or well afterwards, or indeed never? (p. 76)

Ethical engagement with those who are "studied" and "represented" in field research is an important element of the subjective-objective mix Coles finds in documentary work.  These ethical considerations are themselves heightened, for Coles, by the responsibility that individual documentarians have the conduct and products of their craft.

All documentation . . .  is put together by a particular mind whose capacities, interests, values, conjectures, suppositions and presuppositions, whose memories, and, not least, whose talents will come to bear directly or indirectly on what is, finally presented to the world in the form of words, pictures, or even music, or artifacts of one kind or another. (p.87)

Coles also takes pains to acknowledge the contributions of subjects themselves to documentary materials in which they appear.  However, the reference to individual "talent" echoes Coles' pragmatic view of documentary work as a craft, a form of field-based practice in which objective and subjective realities are balanced, integrated and mutually enhanced by individual writers, photographers, film makers, and so on.  This emphasis introduces yet another ethical challenge: reconciling long-term purposes of humanism or social reform with the aesthetic demands of creating artifacts that can be appreciated and understood by others.  As Coles puts it, "Documentary work, then, ultimately becomes, for most of us, documentary writing, documentary photographs, a film, a taped series of folk songs, a collection of children's drawings and paintings: reports of what was encountered for the ears and eyes of others."  (p.99)

Coles locates a third set of ethical issues at the somewhat blurred boundary between documentary work and the political or economic interests of those who might support it.  He reviews the complicating sponsorship of Fortune Magazine for Agee and Evans' study and of the New Left Book Club editors for Orwell's(1958) project in writing The Road to Wigan Pier.  In his discussion of the Farm Security Administration (FSA) photographic survey, he reports Roy Stryker's intent to use photographs quite deliberately to stimulate "a growing consciousness of a kind of life that had hitherto, for many, been safely out of sight, and, moreover, beyond imagining" (p. 156).

To Coles' way of thinking, sponsorship and political interests are complicating but not debilitating challenges to documentary work, and he gives them their due in just those terms.  However, he has less to say about individuals or institutions that might have an interest in keeping things "out of sight," or about the nature and source of inequities that might make Stryker's corrective vision a noble rather than self-serving enterprise.

Clearly Coles has not eschewed this kind of socio-political critique just because it seems more subjective than objective, or because it implicates objective realities that may not be apparent to the subjects he studies.  He takes care, for example, to describe how the value of An American Exodus (1969) depends not only on Lange's evocative photographs but on the hard-nosed economic and political analysis of her husband and collaborator, Paul Schuster Taylor.  In Coles words, Taylor "learned about the individuals, the locale she [Lange] was photographing: how much workers got paid for picking crops, how much they paid for living in a migratory labor camp, and, more broadly, what had happened in the history of American agriculture form earlier years of this century to the late 1930's" (p. 102).  An American Exodus is both a document and a work of social criticism.  But Coles is unclear about whether this kind of criticism should be seen as essential to, or only possible within, the documentary work he values.

Doing Documentary Work includes several verbatim accounts of discussions between Coles and social activists about just this question.  These accounts reveal that Coles' sensibilities overlap only in part with the direct activism he admires in others.  Even when he's interested in redressing wrongs or reducing inequities, Coles is still attending to the dignity of individuals, in an imperfect and yet endearing way.  For the individuals he has met, directly or through the work of the documentarians he values, Coles' inclinations tend towards empathy rather than reproach.  An appreciation of the individual persons he has encountered may also discourage him from casting stones towards those he hasn't - at least not yet.  He notes in passing that while marching in Mississippi to secure dignity for blacks, he and other civil rights activists met with a hostile, at times violent, response from local whites, some of whom might have been children or grandchildren of the same poor whites whose dignity was memorialized by Agee and Evans.  But in this and other instances, Coles offers a well-phrased and thoughtful irony, not a call to arms. 

In his detailed and erudite attention to the artifacts and craft of documentary work, Coles has helped define a significant tradition of practice from which social researchers and others conducting visual studies have a great deal to learn.  The range of examples and issues he examines is both impressive and instructive, and he has written with great insight and experience about ethical and representational matters and the subjective experience of  field work. 

However, while Coles' approach to documentary practice is illuminating, the documentary artifacts he attends to are, by and large, finished, publicly distributed, widely recognized books, photographs, novels, and films.  These works certainly take on new meanings under Coles' documentary gaze.  But this perspective also isolates the promise of "documentary work" from academic disciplines in which more provisional "documents" contribute routinely to social research -- and from a broader conception of science in which work of this sort could play a larger and more visible role. 

Doing Visual Ethnography

In introducing Doing Visual Ethnography, Sara Pink notes that her book "is primarily for ethnographers who wish to incorporate the visual into their ethnographic work, but it is also for photographers and video makers who wish to gain a deeper understanding of how ethnographic research may inform their artistic practice."  Both audiences may indeed profit from a close reading of this text.  Beyond that, Pink refers to blurred boundaries between these two forms of craft as a good thing.  She offers guidelines for moving in that direction, and she illustrates both desired and undesired outcomes with anecdotes and examples from her own field work in Spain and Guinea-Bissau.

Pink has organized the eight chapters of Doing Visual Ethnography into three parts.  "Thinking about visual research" (Part One) includes a review of how visual materials and research strategies have surfaced in various forms of ethnographic research.  It also provides a very good introduction to practical and ethical issues of planning and practicing what Pink calls "visual methods."  Pink's approach runs parallels to Coles' in presenting the work she values as an alternative to conventional scientific inquiry.  But the references Pink cites in explicating a more "reflexive" approach are, by and large, to academic scholarship  -- not to the documentarians, artists, photographers and essayists whose work Coles celebrates in Doing Documentary Work.  This ties Doing Visual Ethnography to a quite different line of discourse and brings to light some analytical issues not addressed directly by Coles. 

One of these is an explicit discussion about the role of subjective and objective realities within scientific inquiry. Coles places the documentary work he values outside the realm of science -- and suggests it should probably stay there.  But Pink wants to extend ethnography -- if not science itself -- to new ways of studying and representing social life, including some that resemble "documentary" works examined by Coles.  

Pink's touchstones for advancing this agenda are largely anthropologists, sociologists, and ethnographic film makers, works by James Clifford (1986), Elizabeth Edwards (1992), and David MacDougall (1997) figure prominently in her arguments.  Pink draws on Clifford's analysis of ethnographic writing as "constructed narratives" to define ethnography as a form of "fiction."  But she also reminds us that Clifford "used the term 'fiction,' not to claim that ethnographies are 'opposed to the truth' or are 'false,' but to emphasize how ethnographies cannot reveal or report on complete or whole accounts of reality; they only ever tell part of the story" (p. 8). 

Taking heart from Clifford's analysis, Pink describes a host of creative strategies for using images to study and represent social life, strategies that go well beyond "recording" human behavior.  These include: giving copies of her own photographs to research subjects and tracking their subsequent distribution through local and extra-local networks; accompanying subjects while they viewed photographs and other images embedded in local cultural events; collaborating with subjects in creating video documents and presentations; examining how particular images are replicated and redefined within and across different cultural settings, and so on. 

These strategies are interesting in their own right, but Pink sees in them a potentially new paradigm for conducting social research.  As she puts it, "by paying attention to images in ethnographic research and representation it is possible that new ways of understanding individuals, cultures and research materials may emerge."  Pink's optimism about this prospect is engaging, as is her advocacy in noting that, "While images should not necessarily replace words as the dominant mode of research or representation, they should be regarded as an equally meaningful element of ethnographic work."  (p 4)

Achieving this kind of parity for words and images, according to Pink, requires a fundamental departure from conventional scientific practice. Towards that end, Pink recommends a form of ethnography in which traditional tenets of "scientific realism" give way to "reflexivity" and "collaboration."  In describing what this means for ethnographers who use photography, she notes the following:

This [reflexive approach] involves: first, developing a consciousness of how ethnographers play their roles as photographers in particular cultural settings, how they frame particular images, and why they choose particular subjects; second, a consideration of how these choices are related to the expectations of both academic disciplines and local visual cultures; and third, an awareness of the theories of representation that inform their photography.  (p. 54)

Pink's recommendation of "reflexive" ethnographic video strikes a similar chord:

The approach I have advocated is critical of the realist stance and of methods texts that limit the potential of video to recording focus group discussions and interviews to avoid "losing" important visual data and cues.  Video is undoubtedly good for such visual 'note-taking,' but such uses ought to be qualified with a rejection of the naive assumption that video records an untainted reality in favour of a reflexive approach that accounts for how video can become part of a focus group discussion or interview. (p. 87)

As these comments illustrate, Pink's use of the term "reflexive" refers to more than one aspect of ethnographic work, and this is both an asset and a liability of Doing Visual Ethnography.  On the one hand, Pink enriches our understanding of field work by examining several dimensions of ethnographic craft to which the word reflexive might apply -- the relationship between images and text, for example, or between researchers and research subjects, prior knowledge and current observations, inquiry and representation, mass mediated and private imagery, friendship and colleagueship, local and extra-local networks, and so on.  Pink is absolutely right to look closely at ethnography in relation to these issues and to challenge the notion that ethnographic practice can be understood apart from them.

That said, Pink refers to so many different interaction effects as "reflexive" that the term begins to loose its punch.  And the contrasts she notes between different interactional forms confounds this all the more.  Are "reflexive" interactions between text and image analogous to those between researchers and subjects, or between prior knowledge and fresh observations?  In some respects they might be, but Pink provides numerous accounts from her own field work experience that illustrate how different sorts of interaction are, indeed, "different". 

The kinds of "reflexive" work that Pink recommends fall within a vision of ethnography that differs in important ways from how traditional anthropologists and sociologists have regarded their craft.  In separating this kind of field work cleanly from "scientific realism," Pink notes that,

Rather than being a method for the collection of 'data,' ethnography is a process of creating and representing knowledge (about society, culture and individuals) that is based on ethnographers' own experiences.  It does not claim to produce an objective of 'truthful' account of reality, but should aim to offer versions of ethnographers' experiences of reality that are as loyal as possible to the context, negotiations and intersubjectivities through which the knowledge was produced. (p. 18)

Consistent with, but not dependent on, this intersubjective notion of ethnographic work, Pink argues for increased community-building among researchers and research subjects.  This ideal is not new to ethnographic work -- see, for example, the essays edited and collected several decades ago by Hymes (1974) and published under the title, Reinventing Anthropology.  However, Pink has the idea that movement in this direction can be facilitated by working with images as well as with words.  As she puts it, "ethnographic photographs can potentially construct continuities between the visual culture of an academic discipline and that of the subjects or collaborators in the research."  

An interest in expanding discourse between researchers and research subjects also leads Pink to an insight of great value for refining ethnographic practice:  That fieldworkers need to consider research materials and strategies -- and ethical guidelines -- in light of the culture in which they will be applied, not just the home culture of the practicing anthropologist or sociologist.  Pink's treatment of this principle is enriched substantially by illustrations from her own field work experience.  As one example, she recounts how still photographs and video tape were regarded quite differently within the local communities she studied in Spain and Guinea-Bissau.  Still photography was a common, widely accepted feature of the Spanish bullfighting sub-culture she was investigating, and this created research opportunities that simply did not exist in Guinea-Bissau.  As she became more aware of these differences, Pink had to develop alternative research strategies and consider quite different ethical issues.  These led her to abandon in Guinea Bissau the still photographic strategies that worked so well in Spain in favor of collaborative video taping.  The latter seemed to work quite well in Guinea Bissau, but it might have been less effective and more problematic, on ethical grounds, in studying Spanish bull fighting culture.

Pink regards the observations and interactions in Spain and Guinea-Bissau that led her to adopt one methodological or ethical posture over another as instances of the "reflexive" and "intersubjective" approach she recommends.  But these can also be regarded as thoughtful efforts by an alert field worker to broaden the range of "realist" data being considered within a given field study.  Indeed, it's difficult to argue that reflexivity on its own negates realist or materialist concerns (Biella, 1988).  And reflexivity of the sort that Pink recommends ironically affirms them by referring to "detailed knowledge of the culture being studied" as a key criterion for assessing, designing or refining field work practice.  Pink rightfully contends that developing knowledge of that sort requires attention to the researcher's own subjectivity.  But it requires something more to separate social analysis from personal fantasies that may bear little if any relation to the culture under study.

The question this raises -- and it corresponds to a parallel question about Coles' work -- is whether rejecting scientific realism is more or less likely to enrich the kind of image-based field work that Pink clearly values.  Pink argues explicitly that advocates of a "realist" approach to image-based field work -- she refers here work by Collier (1967), Grady (1996), Mead (1995), Prosser (1998) and Prosser and Schwartz.(1998) -- are limited in their appreciation of visual materials.  To Pink's way of thinking, those who use photography and video recording solely to "collect data" overlook the value of visual ambiguity, a value better realized in the more "reflexive" work of Banks (1998), Barndt (1997), Chaplin (1994),  and Thomas (1997) .  But creating unfettered ambiguities is not the goal of these other social analyses, nor is ambiguity absent in any intentional form from the inquiries and representations of those whose "realist" sensibilities Pink calls to task. 

In Parts II and III of Doing Visual Ethnography Pink shifts the emphasis of her account from explicit theorizing to strategies for conducting and reporting visual studies.  While some of the strategies she recommends depart from traditional ethnographic practice, they all seem reasonable and appropriate -- not only within Pink's version of "reflexive" ethnography but also within thoughtful field research, broadly defined. 

Part Two includes three very useful chapters about the craft of visual ethnography.  One focuses on photographic applications, one on the use of video, and the third on the challenges of classifying and interpreting visual materials.  Each chapter is organized in a similar way, beginning with an overview of related work and then offering suggestions for "getting started," viewing images made by the researcher, interviewing with images, viewing images made by informants, and so on.  These chapters provide an excellent balance of theoretical musings and practical advice. Pink's field work experience and methodological wisdom shine through them all. 

In examining the strengths and weaknesses of these different research strategies, Pink refers frequently to what she learned from others in the field, from looking closely at situations in which she found herself, and from passing  between one social world and another.  She describes insights that emerged for her when a woman came dressed in fine clothes for what Pink had initially considered an informal photographic encounter.  She recounts how the photograph of a female bullfighter was passed around and commented on by the bull fighter, mutual friends, members of the bull fight media, feminist scholars, and her own academic colleagues.  These observations reveal a keen eye for developing "detailed knowledge" of the cultures Pink studied and her skills in extending such knowledge beyond her immediate personal experience. 

In Part 3, Pink focuses directly on issues of representation.  She examines the relative merits and shortcomings of photographs, printed text, video, and hypermedia.  In addition to thoughtful commentaries about each of these media, Pink makes three general points:  First, because visual imagery is itself always somewhat ambiguous, research reported through such imagery provides distinctive opportunities for readers, audiences, and other witnesses to engage in reflexive inquiry of their own.  Second, non-linear formats for presenting such materials -- such as hypertext, web sites and CDR's -- augment further these opportunities.  And third, this is all to the good. 

In support of the first two arguments, Pink cites examples from her own field work and the work of other scholars, including Barndt (1997), Barbash and Taylor (1997), Berger and Mohr (1967; 1982), Biella (1997) and Stirling (1998 [1965]).  For the third point. Pink re-states her preferences for what she calls "non-hierarchical" representations of knowledge and their value to building open, multi-vocal communities.  In support of this kind of community building, Pink challenges us to consider forms of representation that are less coherent and authoritative than academic scholarship has typically embraced.  Her suggestions about the role of hypertext and multi-media projects are a good case in point.

I suggest that a reflexive approach to knowledge may be applied to hypermedia.  First, by embedding reflexivity in the text itself as an element of hypermedia representation and, secondly, by encouraging hypermedia users to take a reflexive approach to how they create knowledge through their own interaction with hypermedia, thus developing reflexivity as practice.  (p.154)

The constructivist, "reflexive," and democratic ideals to which Pink subscribes are well worth considering further, not only for hypertext formats but for "digital video" as well -- see for example MacDougall's recent essay (2001) and the responses by Ruby, Pink, Wessels, and MacDougall himself in Anthropology Today, 17 (5).  However, because Pink paints "knowledge" with the same broad brush she uses for the term "reflexivity," her argument that "reflexive" materials encourage "reflexive" knowledge is incomplete.  The absence of authorial direction can invite audiences to construct new forms of knowledge, broadly defined, but popular and vernacular forms of knowledge are as likely to denigrate and confound culture and social life as they are to dignify and clarify it. 

Doing Visual Studies

Pink and Coles both champion forms of field work and representation that combine text and images.  They both alert us to ethical complexities that surround this kind of work and speak favorably about collaboration with research subjects.  Both authors also think there's something wrong with a social science that rejects or trivializes visual imagery in studying culture and social life.

But important differences also appear in how these two authors approach their ethical and representational concerns.  In Coles' examination of ethical issues, interpersonal relationships between researcher and research subjects are paramount.  Pink attends to these matters, but she also writes more explicitly about how researchers and their works participate in larger social and economic structures, including mass media enterprises, global networks supporting travel and trade, and so on.  Pink also gives us more to think about than Coles in understanding how the form that ethical issues take can differ substantially in response to social and material differences between photography, video, printed text, multi-media and the World Wide Web. 

Pink and Coles also differ in how they regard the desired outcomes of "ethnographic" or "documentary" work.  For Coles this is a documentary artifact that succeeds or fails in communicating what the documentarian "encountered" for the "ears and eyes of others."  For Pink, however, the documentary artifact appears as a prompt for continuing inquiry with colleagues or with research subjects themselves. Along the same lines, Pink sees great potential in hypermedia for stimulating more broadly based forms of ethnographic inquiry and for supporting communities of researchers, research subjects and other interested parties.  Her interests in that regard resemble the admiration Coles expresses for documentary "works" that can enrich understanding of others and their social circumstances.  But Coles does not consider hypermedia or unstructured documents as a virtue, and the kind of coherence, talent, and personal vision he celebrates in exemplary "documentary work" run counter to Pink's attraction to dynamic, hyper-mediated representations co-authored by researchers and their audiences. 

These parallels and contrasts suggest somewhat different emphases within the critiques Coles and Pink present of contemporary social science.  But what about their shared concerns?  Are these attuned more to inadequacies of social theory or to shortcomings in the practice of individual social scientists?   And if conventional science does fall short as a template for guiding the kinds of work Coles and Pink value, are the "documentary" and "reflexive" frameworks they propose reasonable alternatives?

One challenge to answering these questions is the lack of direct attention Coles and Pink give to the empirical realism that anchors both their accounts. Both authors take great pains to deliberately distance themselves from the realist romance --that empirical, external realities are self-evident and the sole constituent of scientific work and all useful knowledge.  This romance should be laid to rest, and Coles and Pink certainly help that process along.  However, the accounts they provide of "doing" documentary work and visual ethnography also ask us to "look closely" at external realities, learn "about" other people and the circumstances of their lives, and notice the "world" -- not just the contours of an isolated, observing self.  What takes precedence here, what they say about how we should think, or what they say we should do?

Another challenge to answering these questions is that few social scientists have the experience, appreciation and technical skills that Coles and Pink demonstrate for working with the varied media and genres they value so highly.  This is particularly true when it comes to representing observations and research findings in ways that are aesthetically coherent and accessible to lay audiences.  It's not always clear what social scientists would do with photography, film and vivid narrative texts if they had greater skills and training in work of that kind.  With this in mind, at least part of the problem that Coles and Pink have with "social science" may be that the training of social scientists neglects this kind of representational craft.

But what about the issue of social theory itself?  Are the forms of "visual study" that Coles and Pink recommend defensible or not within the logic of social science?

This is not a question that either Coles or Pink invite us to ask.  Each argues for protecting their own version of image-based field work from the distractions and compromise of trying to defend it in purely "scientific" terms.  Trying to fit documentary work or visual ethnography into realist social science, they argue, would distort their central purposes and diminish their contributions to social and cultural understanding.

I respect such sentiments.  However, there are ways for advancing the kind of work that Coles and Pink recommend as thoughtful forms of realist scientific practice -- without the distortions and compromise they fear.  To understand the correspondence that could lead in this direction, consider the following five propositions: 

(1) The most essential feature of realist scientific practice is a commitment to test ideas against what appear to be relevant and telling external realities. 

(2) Practitioners of empirical, scientific inquiry have found myriad ways to perform these tests and to describe their results, some of which have been codified into conventions of scientific practice by different communities of inquiry.

(3) Apart from the craft skills that individual develop in conducting and reporting such tests -- and the social organization of institutions that support and take advantage of this craft -- the practice of science is embedded in the larger culture and continuous with other forms of social life. 

(4) As a result of this embeddedness and continuity, scientists will always be vulnerable -- and reasonably so -- to challenges that their work is shaped by their own self-interest,  social position, and subjectivity. 

(5) These challenges stimulate scientists to develop a secondary set of crafts for explaining their work to others in terms that acknowledge and distinguish between external realities and their own subjectivity. 

Coles and Pink see notions similar to propositions 3 and 4 as integral to documentary work and visual ethnography and as features that set such work apart from the conventions of realist social science.  They also define the latter almost exclusively in terms of proposition 1 and 2.  However, their detailed accounts of "doing" documentary work and visual ethnography implicitly affirm all five propositions, and much the same can be said for recent scholarship on the social organization of scientific practice.  This is not to say that all of what Pink and Coles have to say about documentary work or visual ethnography bears repeating as observations about the practice of "realist social science."  But it's hard to argue that the portrayals Pink and Coles provide of these two kinds of "doing" fall categorically outside the realm of empirical, scientific inquiry, as defined by these five propositions.

None of this is likely to convince scientists already committed to other conventions for "testing empirical reality" to suddenly embrace documentary work and visual ethnography.  And one significant obstacle to that embrace -- whether or not Coles and Pink would welcome it -- is found in the ambiguities of visual imagery noted earlier.  But two other obstacles may actually be more consequential. 

One of these refers to the second proposition noted above -- the craft of designing and describing how ideas are tested against empirical realities.  Both Coles and Pink affirm this craft in the examples they cite of exemplary field work practice.  However, at just the point where a realist social scientist might look for "systematic attention to varied data sources," or "authentication strategies," or "criteria for assessing validity," Coles and Pink disavow realist science itself.  This invites thoughtful readers, whether or not they are scientists, to wonder how Coles and Pink might distinguish the best of social observation and inquiry -- in the form of images, text, multi-media or film or even novels -- from the worst of speculative social theorizing.  To overcome this obstacle we need more explicit strategies than either author provides for assessing the evidentiary foundations of image-based field work --varied and imperfect though these may be.

Another obstacle to the scientific embrace of documentary work and visual ethnography has less to do with what Coles and Pink provide or don't provide than with additive and de-contextualized ideals for "creating new knowledge" -- ideals to which science is closely aligned by scientists and lay people alike.  The notion that science proceeds in a social vacuum by accumulating bits and pieces of denatured and incontrovertible empirical data is very attractive.  Indeed, focusing entirely on the first two propositions noted above offers both scientists and citizens the comfort of ignoring uncertainties, complications and ambiguities suggested by the three that follow.

It's a false comfort.  Kuhn (1962) noted some decades ago that science proceeds not only by accumulating bits of empirical evidence, but also through contests between one set of ideas and another.  This observation has since been confirmed and refined by studies of the culture of scientific practice that follow a trajectory defined by Latour and Woolgar (1986) and other social researchers.  The "contests" examined in most social studies of physical and natural science are defined in terms that have much greater meaning for scientists themselves than for the larger public, but in the social sciences, boundaries between expert and lay versions of these contests are frequently blurred.  This creates for even the most limited "data" elements, a potential aura of great social significance.  A transcript of a single interview can challenge a broadly applied social theory.  A video tape recording of parent and infant interaction can confound lay and expert ideas about language acquisition. Photographs of buildings taken over 30 years time can confirm or disconfirm ideas about social change that shape both research and public policy. 

The single transcript, video tape, or photograph, of course, will never provide evidence enough to construct from scratch comprehensive social theories.  But that's a bit beside the point, because starting from scratch in enterprises of this sort is not really an option (nor can anyone reserve the last turn in developing social theory).  Just as our best knowledge of scientific practice is consistent with all five of the propositions noted above, not just the first two, so too the craft it requires is more akin to "reconstructing" theories than constructing them (Burawoy, 1991). 

This generates for the social sciences both the opportunity and the need for public forms of inquiry.  As a "reconstructive" enterprise, a social science that supports this kind of inquiry could be nourished by dispassionate collection of so-called "objective" data through various forms of social accounting.  But it can also benefit from investigations that follow, in one way or another, Stryker's interest in stimulating a "consciousness" of things that were otherwise out of sight or "beyond imagining."  And it not only benefits from but requires field studies that "document" the relative truth or falsehood of already formed social and cultural attributions, including those embedded in the priorities and technology of social accounting itself. 

The kind of knowledge valuable to such reconstructive work is typically partial, thoughtful, and, at least partly accessible to popular understanding.  The most promising approaches to generating knowledge of that sort reflect broad principles of scientific inquiry, well-informed by propositions such as those outlined above, but they also depend on scholars and other investigators who are expert practitioners of the representational crafts. 

With all this in mind, separating documentary work and visual ethnography from a broader domain of social scientific inquiry has several unfortunate implications. It can reduce the responsibility that practitioners of these two kinds of "doing" feel for clarifying the evidentiary roots of their work.  It can also encourage the misguided notion that working with visual materials -- or other forms of representation that have both evocative and denotative features -- in itself represents a step away from scientific concerns (though there's plenty of recent evidence to the contrary).3 Pulling documentary work and visual ethnography too far away from science can also pull social scientists themselves away from representational crafts through which their work might contribute more to public awareness and debate. 

In collecting and making images and other artifacts, and using these to describe to others the cultural and social realities we care about, we can set about to "collect data," or to deliberately challenge or confirm what people already know to be true, or to present artifacts as a window into our own subjectivity, or to offer them up as evidence of other subjective realities, including those tied to negotiated inquiry with people we have studied.  The fact that whatever visual materials we examine or produce are always all of these things is neither cause for despair nor a license for groundless speculation.  It's the very foundation of those crafts Coles and Pink invite us to explore further, seeking truths and avoiding error in whatever ways we can.


Notes

1 It's useful to remember that "traditional" scholarship was a form of classical inquiry grounded in the interpretation of texts.  In more democratic communities, the texts lent themselves to multiple interpretations; in less democratic ones, they did not.  The move towards "modern" scholarship was a move towards a plurality of texts and towards assessing texts and speculation through empirical evidence.  Empiricism thus emerged as an alternative and corrective to speculation and textual analysis.  Ironically, by problematizing empirical alternatives to speculation and textual analysis, some forms of post-modern scholarship recapitulate both the strengths and weaknesses of classical studies. 

2  Coles' recent engagements with documentary work have included a close association with the Center for Documentary Studies at Duke University and his editorship of Doubletake magazine.

3 In June, 2001, the National Academy of Sciences issued a report recommending strongly that increased support be provided for video tape studies of schools and classrooms.  Some recent examples of legitimated visual study that range from "data collection" to "reflexive inquiry" include: the video component of the multi-nation Third International Mathematics and Science Study (TIMSS) (Stigler, Gonzales, Kawanka, Knoll & Serrano, 1999), an iterative, community building study by Tobin, Wu and Davidson of pre-schools in three cultures (Tobin, Wu & Davidson, 1989); Anne Dyson's microethnographic studies of how children use text, speech and drawings in negotiating classroom cultures (Dyson, 1989; 1993; 1997); and studies by Rayner and others based on visual evidence of eye movements in reading (Rayner, 1998).


References

Agee, J., & Evans, W. (1960 [1939]). Let Us Now Praise Famous Men. Boston: Houghton-Mifflin.

Banks, M. (1998). Visual anthropology: Image, object and interpretation. In J. Prosser (Ed.), Image-Based Research, (pp. 9-23). London: Taylor and Francis.

Barbash, I., & Taylor, L. (1997). Cross-Cultural Filmmaking: A Handbook for Making Documentary and Ethnographic Films and Videos. Berkeley, CA: University of California.

Barndt, D. (1997). Zooming out/zooming in: Visualizing globalization. Visual Sociology, 12 (2), 5-32.

Berger, J., & Mohr, J. (1967). A Fortunate Man. Cambridge: Granta.

Berger, J., & Mohr, J. (1982). Another Way of Telling. Cambridge: Granta.

Biella, P. (1988). Against reductionism and idealist self-reflexivity: The Ilparakuyo Maasai Film Project. In J. R. Rollwagen (Ed.), Anthropological Filmmaking, (pp. pp. 47-72). New York: Harwood.

Biella, P. (1997). Mama Kone's possession: scene form an interactive ethnography. Visual Anthropology Review, 13(1), 59-95.

Burawoy, M. (1991). Ethnography Unbound. Berkeley, CA: University of California.

Chaplin, E. (1994). Sociology and Visual Representations. London: Routledge.

Clifford, J., & Marcus, G. E. (Eds.). (1986). Writing Culture: The Poetics and Politics of Ethnography. Berkeley, CA: University of California.

Coles, R. (1992). Their Eyes Meeting the World: The Drawings and Paintings of Children. New York: Houghton Mifflin.

Coles, R. (1997). Doing Documentary Work. New York: Oxford University.

Coles, R., & Nixon, N. (1998). School. Boston, MA: Little, Brown.

Collier, J. J. (1967). Visual Anthropology: Photography as a Research Method. New York: Holt, Rinehart & Winston.

Dyson, A. H. (1989). Multiple Worlds of Child Writers: Friends Learning to Write. New York: Teachers College Press.

Dyson, A. H. (1993). Social Worlds of Children Learning to Write in an Urban Primary School. New York: Teachers College.

Dyson, A. H. (1997). Writing Superheroes: Contemporary Childhood, Popular Culture, and Classroom Literacy. New York: Teachers College.

Edwards, E. (1992). Anthropology and Photography. New Haven, CT: Yale University.

Grady, J. (1996). The scope of visual sociology. Visual Sociology, 11(2), 10-24.

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

Kuhn, T. S. (1962). The Structure of Scientific Revolutions. Chicago: University of Chicago.

Lange, D., & Taylor, P. S. (1969). An American Exodus. New Haven: Yale University.

Latour, B., & Woolgar, S. (1986). Laboratory Life: The Construction of Scientific Facts. Princeton, NJ: Princeton University Press.

MacDougall, D. (1997). The visual in anthropology. In M. Banks & H. Morphy (Eds.), Rethinking Visual Anthropology, . London: New Haven Press.

MacDougall, D. (2001). Renewing Ethnographic Film: Is digital video changing the genre? Anthropology Today, 17(3), 15-21.

Mead, M. (1995). Visual anthropology in a discipline of words. In P. Hockings (Ed.), Principles of Visual Anthropology, (pp. 3-10). New York: Mouton de Gruyter.

Orwell, G. (1958). The Road to Wigan Pier. New York: Harcourt, Brace.

Pink, S. (2001). Doing Visual Ethnography. London: SAGE.

Prosser, J. (Ed.). (1998). Image-Based Research: A Sourcebook for Qualitative Researchers. Bristol, PA: Falmer Press.

Prosser, J., & Schwartz, D. (1998). Photographs within the sociological research process. In J. Prosser (Ed.), Image-Based Research, (pp. 115-130). London: Taylor and Francis.

Rayner, K. (1998). Eye movements in reading and information processing: Twenty years of research. Psychological Bulletin, 124, 372-422.

Stigler, J. W., Gonzales, P. A., Kawanka, T., Knoll, S., & Serrano, A. (1999). The TIMSS Videotape Classroom Study: Methods and Findings from an Exploratory Research Project on Eighth-Grade Mathematics Instruction in Germany, Japan, and the United States (Research and Development Report: Third International Mathematics and Science Study (TIMSS) NCES 1999074). Washington, DC: National Center for Education Statistics.

Stirling, P. (1998 [1965]). Turkish Village. http://lucy.ukc.ac.uk.TVillage/StirlingContents.html

Thomas, H. (1997). Dancing: Representation and Difference. In J. McGuigan (Ed.), Cultural Methodologies, . London: SAGE.

Tobin, J. J., Wu, D. Y. H., & Davidson, D. H. (1989). Preschool in Three Cultures. New Haven: Yale University Press.