Visual Sociology: Photography and Sociology

PHOTOGRAPHY AND SOCIOLOGY

HOWARD S. BECKER
Northwestern University

Photography and sociology have approximately the same birth date, if you count sociology's birth as the publication of Comte's work which gave it its name, and photography's birth as the date in 1839 when Daguerre made public his method for fixing an image on a metal plate 2. From the beginning, both worked on a variety of projects. Among these, for both, was the exploration of society.

While sociology has had other ends, moral and metaphysical, sociologists have always wanted to understand how society worked, to map its dimensions and then look into the big sectors and little crannies so mapped. They ordinarily wanted to find things out rigorously and scientifically, and to develop general theories. But some sociologists have made it their main business to describe what has not yet been described, in the style of the ethnographer, to tell the big news, in the style of the journalist, combining these (more or less) with the desire for rigor and general theory.

Sociologists' choice of theories, methods, and topics of research usually reflect the interests and constraints of the intellectual and occupational communities to which they are allied and attached. They often choose research methods, for instance, that appear to have paid off for the natural sciences. They frequently choose research topics which are public concerns of the moment, especially as those are reflected in the allocation of research funds: poverty, drugs, immigration, campus or ghetto disorder, and so on. These faddish tendencies are balanced by a continuing attention to, and respect for, traditional topics and styles of work.

The efforts and projects of photographers have been much more various. In order to understand how photographers go about exploring society when they undertake that job, it will be useful to remember the melange of other jobs photography does. Think of a camera as a machine that records and communicates much as a typewriter does. People use typewriters to do a million different jobs: to write ad copy designed to sell goods, to write newspaper stories, short stories, instruction booklets, lyric poems, biographies and autobiographies, history, scientific papers, letters . . . . The neutral typewriter will do any of these things as well as the skill of its user permits. Because of the persistent myth that the camera simply records whatever is in front of it (about which I will say more below), people often fail to realize that the camera is equally at the disposal of a skilled practitioner and can do any of the above things, in its own way. Photographers have done all of the things suggested above, often in explicit analogue with the verbal model. Different kinds of photographers work in different institutional settings and occupational communities, which affect their product as the institutional settings in which sociologists work affect theirs (Rosenblum 1973).

Photographers have worked to produce advertising illustrations. They have made portraits of the rich and famous, and of ordinary people as well. They have produced pictures for newspapers and magazines. They have produced works of art for galleries, collectors and museums. The constraints of the settings in which they did their work (Becker 1974) affected how they went about it, their habits of seeing, the pictures they made and, when they looked at society, what they saw, what they made of it and the way they presented their results.

From its beginnings, photography has been used as a tool for the exploration of society, and photographers have taken that as one of their tasks. At first, some photographers used the camera to record far-off societies that their contemporaries would otherwise never see and, later, aspects of their own society their contemporaries had no wish to see. Sometimes they even conceived of what they were doing as sociology, especially around the turn of the century when sociologists and photographers agreed on the necessity of exposing the evils of society through words and pictures. Lewis Hine, for instance, was supported by the Russell Sage Foundation in connection with the early surveys of urban life (Gutman 1967). The American Journal of Sociology routinely ran photographs in connection with its muckraking reformist articles for at least the first fifteen years of its existence (Oberschall 1972:215).

Another kind of social exploration grew out of the use of photographs to report the news and to record important social events. Mathew Brady (Horan 1955) and his staff, which included Timothy H. O'Sullivan (Horan 1966) and Alexander Gardner (1959), photographed the Civil War, and Roger Fenton the Crimean War. But it was not until the 1920's that the development of the illustrated weekly in Europe produced a group of photographers who made the photoreportage or photoessay into an instrument of social analysis (Alfred Eisenstaedt and Erich Salomon are among the best-known graduates of these journals) (Gidal 1973). Later, the Picture Post in England and Time, Life, and Fortune in the United States provided outlets for serious photojournalists who worked with the photoessay form: Margaret Bourke-White, Walker Evans, W. Eugene Smith, Robert Capa.

The impulse to photographic social exploration found another expression in the work produced by the photographers Roy Stryker assembled for the photographic unit of the Farm Security Administration during the 1930's (Hurley 1972, 1973; Stryker and Wood 1973). Dorothea Lange, Walker Evans, Russell Lee, Arthur Rothstein, and others made it their business to record the poverty and hard times of Depression America, their work very much informed by social science theories of various kinds.

More recently, political involvement has had a hand in shaping the use of photography to explore society.

Photographers participated actively in the civil rights movement of the 1960's and brought back photographs which effectively stirred people just as Hine's photographs of child laborers had. They then used those skills in somewhat less immediately political kinds of essays-exploring communities, occupations, subcultures, institutions-that have a sociological intent. These essays combine a journalistic and ethnographic style with a self-conscious and deliberate artistic purpose.
Photography from the beginning strove toward art just as it did toward social exploration. To be sure, earlier photographers in this tradition understood that what they did had an artistic component. They worked hard to produce images that measured up as art. But the artistic element of photography was held at a substantial distance from photography carried on for more mundane purposes, including journalism. Such influential photographers as Edward Weston conceived of their work as something more like painting-they produced for galleries, museums, and private collectors as much as they could-and did very little that could be interpreted in any direct way as an exploration of society.

Art and social exploration describe two ways of working, not two kinds of photographers. Many photographers do both kinds of work in the course of their careers. And even this is an over-simplification since many photographs made by someone whose work is predominantly of one kind have strong overtones of the other. Paul Strand is clearly an art photographer; but his pictures of peasants around the world embody political ideas, and any number of socially concerned photographers do work that is personally expressive and aesthetically interesting quite apart from its subject matter - as, for instance, in Danny Lyon's The Destruction of Photography has thus, like sociology, displayed a shifting variety of characteristic emphases, depending on the currents of interest in the worlds of art, commerce and journalism to which it has been attached. One continuing emphasis has been the exploration of society in ways more or less connected with somewhat similar explorations undertaken by academic sociologists. As sociology became more scientific and less openly political, photography became more personal, more artistic, and continued to be engaged politically. Not surprisingly, then, the two modes of social exploration have ceased to have very much to do with one another.

Sociologists today know little of the work of social documentary photographers and its relevance to what they do. They seldom use photographs as a way of gathering, recording, or presenting data and conclusions. I want to acquaint them with this tradition and show them how they can make use of the styles of work and techniques common in photography. Many social scientists have already been active photographically, and what I say will not be news to them (Barndt 1974).

Many photographers have undertaken projects which produce results that parallel those of sociology, and make claims that in some ways parallel the claims to truth and representativeness of sociology. Insofar as their work has this character, I intend to show them how a knowledge of some of the ideas and techniques of academic sociology can be of help to them.

I do not want to make photographers of social scientists or impose a social science imperialism on photographers (not that there is any chance such attempts would be successful). Many sociologists will find the work and methods I describe hopelessly unscientific, although I hope that this discussion will cause them to reconsider their own methods. Many photographers will find my suggestions academically arrogant; satisfied with the way they now work, they will see no advantage in alien ideas and procedures.

What I say is most directly addressed to those social scientists and photographers who are sufficiently dissatisfied with what they are doing to want to try something new, who find difficulties in their present procedures and are interested in seeing whether people in other fields know something that might help. Ideally, it is directed to the growing number of people, whatever their professional background, who are concerned with producing photographic explorations of society.

In addition, I have tried to show how even those sociologists who have no interest in photographic work can learn something from the light shed on conventional research methods by a comparison with photographic methods. Some generic problems of social exploration profit from the light the comparison generates.

I will not be concerned with every aspect of the use of visual materials in social science in this paper. Specifically, I will not consider three major areas of work to which social scientists have devoted themselves: (1) the use of film to preserve nonverbal data for later analysis, as in the analyses of gesture and body movement by such scholars as Birdwhistell, Ekman, Hall, and Lennard; (2) the analysis of the visual productions of "native seers" for their cultural and social meanings, as in the Worth-Adair (1972) study of Navaho filmmakers; (3) the use of photographs as historical documents, whether they have been taken by artless amateurs and preserved in family albums, as in Richard Chalfen's work, or by professional photographers, as in Lesy's Wisconsin Death Trip (1973). All three are interesting and important areas of work, but differ from the use of photographs to study organizations, institutions, and communities that I have in mind. There is considerable overlap, of course, and I do not insist on the distinction.

Anyone who gets into a new field must pay some dues. Photographers who want to pursue the matter further will have to read some social science prose, and many will probably find that too steep a price; some will find a viable solution in a working partnership with a social scientist (as in the fruitful collaboration of Euan Duff and Dennis Marsden in an as yet unpublished study of unemployed men and their families in Britain).

The price to social scientists is less painful. They must acquaint themselves with the extensive photographic literature; I have reproduced some examples here and will provide a brief guide to more. In addition, they will have to learn to look at photographs more attentively than they ordinarily do. Laymen learn to read photographs the way they do headlines, skipping over them quickly to get the gist of what is being said. Photographers, on the other hand, study them with the care and attention to detail one might give to a difficult scientific paper or a complicated poem. Every part of the photographic image carries some information that contributes to its total statement; the viewer's responsibility is to see, in the most literal way, everything that is there and respond to it. To put it another way, the statement the image makes-not just what it shows you, but the mood, moral evaluation and causal connections it suggests-is built up from those details. A proper "reading" of a photograph sees and responds to them consciously.

Photographers learn to interpret photographs in that technical way because they want to understand and use that "language" themselves (just as musicians learn a more technical musical language than the layman needs). Social scientists who want to work with visual materials will have to learn to approach them in this more studious and time-consuming way. The following exercise, taught to me by Philip Perkis, is a way of seeing what is involved:

Take some genuinely good picture; the ones reproduced in this article will do. Using a watch with a second hand, look at the photograph intently for two minutes. Don't stare and thus stop looking; look actively. It will be hard to do, and you'll find it useful to take up the time by naming everything in the picture to yourself: this is a man, this is his arm, this is the finger on his hand, this is the shadow his hand makes, this is the cloth of his sleeve , and so on. Once you have done this for two minutes, build it up to five, following the naming of things with a period of fantasy, telling yourself a story about the people and things in the picture. The story needn't be true; it's just a device for externalizing and making clear to yourself the emotion and mood the picture has evoked, both part of its statement


When you have done this exercise many times, a more careful way of looking will become habitual. Two things result. You will realize that ordinarily you have not consciously seen most of what is in an image even though you have been responding to it You will also find that you can now remember the photographs you have studied much as you can remember a book you have taken careful notes on. They become part of a mental collection available for further work. (When you do this exercise a number of times you will acquire new habits of seeing and won't have to spend as much time looking at a new print).

I hope this does not sound mystical. Black and white still photographs use visual conventions that everyone brought up in a world of illustrated newspapers and magazines learns just as they learn to talk. We are not ordinarily aware of the grammar and syntax of these conventions, though we use them, just as we may not know the grammar and syntax of our verbal language though we speak and understand it. We can learn that language through study and analysis, just as we can learn to understand music and poetry by making technical analyses of harmony and counterpoint or of prosody. We don't have a large amount of such photographic analysis available, especially as it relates to the concerns of social scientists. But it is absolutely prerequisite to any analysis and discussion that you practice looking at photographs long and hard, so that you have something to analyze.

THE PHOTOGRAPHIC LITERATURE

Topics of Study

One reason sociologists should be interested in the work of social documentary photographers is that photographers have covered many of the subjects that are persistent foci of sociological concern. Some have done their work for the government, some on assignment, or speculatively, for magazines and newspapers, some supported by foundations, some as the "private" work they do between paying jobs, or as a hobby. Describing the variety of topics photographers share with sociologists will provide the opportunity to acquaint those unfamiliar with the photographic literature with some of the most interesting and important work.

In dealing with the topics they share with sociologists, photographers say what they have to say in many ways. Without giving many examples, or offering an extended description of the various forms of photographic statements, I'll simply suggest the following as among the possibilities now in use. A photographer may make his statement in the form of an aphorism or witticism, a photographic one-liner that may be no more than a joke (in the case of EIliot Erwitt 1972, for example) or may be of considerable depth (as in the work of André Kertész 1972). He may produce slogans. He may be saying "Look at that!" in wonder at some natural phenomenon (Ansel Adams' pictures of Yosemite seem to say that), or in revulsion from some disgusting work of man (McCullin 1973). He may tell a story or, finally, he may produce something that implicitly or explicitly offers an analysis of a person, an artifact, an activity or a society. It stretches ordinary usage to speak of these projects as "studies," as though they were sociological research projects; but the exaggeration emphasizes, as I want to, the continuity between the two kinds of work.

Both photographers and sociologists have described communities. There is nothing in photography quite like such major works of social science as Warner's Yankee City Series, Lynd 's Middletown and Middletown in Transistion , and Hughes' French Canada in Transition . Photographers have recently produced more modest efforts, such as Bill Owens' Suburbia (1973) and George Tice's Paterson (1972), both describing smaller communities through a hundred or so images of buildings, houses, natural features, public scenes and (in Owens' book) family life. A number of photographers have accumulated massive numbers of negatives of one city, as Eugéne Atget (Abbott 1964) did in his attempt to record all of Paris or Berenice Abbott (1973) or Weegee (1945), the great news photographer, did, each in their way, of New York; but only small selections from the larger body of work are available, and we usually see only a few of the images at a time.

Like sociologists, photographers have been interested in contemporary social problems: immigration, poverty, race social unrest. In that great photographic tradition, one typically describes in order to expose evils and call for action to correct them. Lewis Hine, who called himself a sociologist, put credo succinctly: "I want to photograph what needs to be appreciated; I want to photograph what needs to be corrected." His greatest project showed conditions of child labor in the United states in a way that is thought to have helped the passage of remedial legislation. Somewhat earlier, Jacob Riis (1971), a reporter, photographed the slums of New York and exhibited the results in How the Other Half Lives. I have already mentioned the photographs of rural poverty by the members of Stryker's FSA photographic unit and might add to that the collaboration of Bourke-White and Erskine Caldwell (1937) in You Have Seen Their Faces. Life in Black ghettoes has been photographed, from the inside, by men like James Van Der Zee (DeCock and MeGhee 1973) (among other things the official photographer for Marcus Garvey) and Roy de Carava (de Carava and Hughes 1967); from the outside, by Bruce Davidson (1970) and many others. Dramatic confrontations of the races make news, and many photographers have covered such stories (Hansberry 1964) and gone on to more extended explorations of the matter. W. Eugene Smith (1974) has recently published a major essay on pollution, its victims, and the politics surrounding it in Japan.


Other photographic work deals with less controversial problems, in the style of the sociological ethnography. Sociologists have studied occupations and the related institutions of work, and photographers have too: Smith (1969) did major essays on a country doctor and a Black midwife; Wendy Snyder (1970) has a book on Boston's produce market, and Geoff Winningham (1971) produced a book-length study of professional wrestling. Photographers have also investigated social movements, as in Paul Fusco's (1970) book on Cesar Chavez and the UFW, Marion Palfi's (1973) work on civil rights, or Smith's classic essay on the Ku Klux Klan (1969). They have shared with sociologists an interest in exotic subcultures: Danny Lyon's (1968) work on
motorcycle gangs and Brassai's (Museum of Modern Art 1968) studies of the Parisian demi-monde, for instance.


Photographers have been as alert as sociologists and cultural commentators to call attention to the rise of new social classes or to forgotten groups in society. Two recent books try to do this, using Detroit as the laboratory. Alwyn Scott Turner's (1970) Photographs of the Detroit People concentrates on the working class, in front of their homes in the parks, streets and churches, at parades and rallies. Enrico Natali's (1972) New American People does something similar for the rising middle class.
vMany photographers have worked at depicting the ambiance of urban life in a way reminiscent of the long tradition of theorizing about cities by sociologists from Simmel to Goffman. Walker Evans' (1966) Many Are Called consists of portraits made on the New York subway with a hidden camera. Lee Friedlander, Garry Winogrand (Davidson et al. 1966) and a host of others have photographed "behavior in public places," creating in the mood of their images a sense of alienation and strain, maybe even a little anomie. Euan Duff's (1971) How We Are systematically covers major aspects of urban British life.

In addition to these relatively conventional analogues of sociological investigation, photographers have also been concerned with the discovery of cultural themes, modal personalities, social types, and the ambiance of characteristic social situations. Thus, Robert Frank's (1969) enormously influential The Americans is in ways reminiscent both of Tocqueville's analysis of American institutions and of the analysis of cultural themes by Margaret Mead and Ruth Benedict. Frank presents photographs made in scattered places around the country, returning again and again to such themes as the flag, the automobile, race, restaurants - eventually turning those artifacts, by the weight of the associations in which he embeds them, into profound and meaningful symbols of American culture.

The long tradition of the photographic portrait has led photographers to attempt, in a way sociologists have seldom tried (despite the tradition of the life history document), to depict societies and cultures by portraits of representative types. The most systematic attempt must be August Sander's Men Without Masks, which characterizes Germany in hundreds of portraits of Germans of every social class, occupation, ethnic, regional, and religious group. Paul Strand's (1971) portraits of peasants from France, Egypt, Ghana, Morocco, Canada, and elsewhere, though surrounded by other images of places and artifacts, attempt the same thing, as to Elaine Mayes' (1970) portraits from the Haight-Ashbury.

Photographers have seldom, constrained as they are by time limitations built into the institutions they work in, attempted longitudinal studies. One recent project of this kind suggests how it can happen. Larry Clark's (1971) Tulsa tells the story of a group of young men in that city who begin using intravenous amphetamine. It follows them from an idyllic hunting and fishing youth into drugs, police trouble, and death. Clark was one of the group and visited his old friends periodically as the story unfolded, thus producing a unique inside view of an exotic subculture.
Photographers like to capsulize their understanding of people, situations, even countries, in one compelling image.


Cartier-Bresson (1952) coined the phrase "the decisive moment" to refer to that moment when things fall into place in the viewfinder in such a way as to tell the story just right. It sounds mystical, but many of his pictures (e.g., "Exposing a stool pigeon for the Gestapo in a displaced persons camp", Dessau 1945) accomplish just that.
Modes of Presentation
Photographers present the results of their explorations of society in a variety of ways, using varying quantities of images to make different kinds of statements. One might, at one extreme, present a single image, capturing in it all that needs to be shown about something from some point of view. Stieglitz' "The Steerage," for instance, seems to make a self-sufficient statement about the experience of European immigrants, showing both the masses Emma Lazarus wrote about, crowded onto the deck of the ship, but also a brilliantly lit gangway that seems to lead to better things. (lronically, the ship was actually headed east, to Europe.)
Usually, however, photographers exploring society give us more than one striking image. They explore a topic more thoroughly, sometimes in one concentrated burst of attention and activity, sometimes (on a timetable more like that of the social scientist) over a period of a few years, sometimes as the preoccupation of a lifetime. The concentrated burst occurs when the conditions of work-a magazine assignment, for instance-make it unlikely that you will be able to return to the subject again 4. It may occur when circumstances make a brief visit possible to an ordinarily inaccessible place (Bourke-White's visit to Russia). Photographers can seldom get the support for more long-term projects, certainly not on a routine basis, so a great deal of important work has been done in this concentrated way and many prized photographic skills consist of doing good work despite the lack of sufficient time.

Probably because of the connection with magazine work, such photographic studies typically saw publication as a photoessay. The form, pioneered in Europe, reached maturity in Fortune and Life. Bourke-White, Smith, and others developed a form in which a few to as many as thirty photographs, spread with an accompanying text over four to eight or ten pages, explored a subject in some detail, giving more space and attention to a subject than a conventional journalistic treatment allowed. Photoessays often, like good sociological studies, showed the great variety of people and situations involved in the subject under study. Of course, magazine editors played a decisive part in the selection and arrangement of the materials, and photographers frequently objected to their interference. Gene Smith resigned from Life over this issue.

When a photographer finds it possible to pursue a subject for a longer time-a year or more-he may accumulate sufficient material for a more extended presentation. Guggenheim grants and other fellowship and foundation funds have supported many such projects (Bruce Davidson's East 100th Street, many of Marion Palfi's studies, Smith's work on Pittsburgh). The government has supported others: the FSA projects, Hine's exposes of child labor. Or the project may be the photographer's private affair, supported by work of an entirely different kind.

In any event, photographers who work over a more extended period accumulate a large pool of images from which they can choose those that best express their understanding of their topic. Choices are made from that pool of images for specific uses, often in consultation with or entirely by others: editors, curators and the like. The selection so made may have more or less organization and coherence. The work of the FSA photographers, for instance, typically appears simply as a collection of variable size and made up of a variety of combinations from the entire body of work they produced.

Larger selections of work usually appear either as books, museum exhibits or both. They may contain anywhere from thirty to four or five hundred prints. Especially when they appear as books, the projects often take on a more organized and sequential format. Such formats allow, and almost require, a more analytic stance than a simple collection, and suggest statements that overlap considerably with those found in sociological ethnography.
The function of text in a photographic book is not clear. Photographic books may contain no text at all (e.g., Davidson's East 100th Street). In others, photographs are presented with a brief identifying label, often no more than a place and date, as in Frank's The Americans. Some contain a paragraph or so of commentary on many of the images, as in Leonard Freed's (1970) Made in Germany. Still others contain large chunks of independent text-as in Danny Lyon's Bikeriders (1968) or Conversations with the Dead (1971) or Winningham's studies of wrestlers (1971) and rodeos (1972)-taken from extant documents or tape-recorded interviews. Finally, as in Smith's essay on pollution in Minamata, the photographer may include an extensive explanatory and analytic text

THEORY IN PHOTOGRAPHY

Close study of the work of social documentary photographers provokes a double reaction. At first, you find that they call attention to a wealth of detail from which an interested sociologist could develop useful ideas about whose meaning he could spin interesting speculations. A collection of photographs on the same topic-a photographic essay or book-seems to explore the subject completely. Greater familiarity leads to a scaling down of admiration. While the photographs do have those virtues, they also tend to restrict themselves to a few reiterated simple statements. Rhetorically important as a strategy of proof, the repetition leads to work that is intellectually and analytically thin.

Many sociologists and photographers will find those judgments irrelevant. Some sociologists work with equally simple ideas; but those who are responsive to the tradition of ethnographic fieldwork will want photographic explorations to provide results as rich and interesting as their own descriptions. Some photographers are content to produce a few compelling images. But many of the book-length projects just described aspire to more than that, whether they make the aspiration explicit or not. Their authors are sensitive to the currents of thought and interest in the larger cultural community, and want to do work that is thought of as more than a beautiful illustration. Photographers and sociologists who don't share these traditions and sensitivities will find what follows of little use.

The problem, then, is why photographic exploration of society is so often intellectually thin. A subsidiary question of interest to photographers and to sociologists who may take a photographic approach to their work, is: what can be done to make that work intellectually denser?

The answer to these questions lies in understanding the role of theory in making photographs of social phenomena. Most sociologists accept the folk notion that the camera records objectively what is there for it to record, no matter what the ideas of the person who pushes the button. Laymen may believe this, but photographers know better. To be sure, something real has to emit light rays in order to produce an image on film or paper, and whatever is real that is emitting light rays where they can go through the lens will make some kind of image. That constraint exists, so that John Collier, Jr. (Friends of Photography 1972:49) is right to say that "The camera constantly trips up the artist by loyally going on being a recorder of reality."

Nevertheless, the photographer exerts enormous control over the final image and the information and message it contains. The choice of film, development and paper, of lens and camera, of exposure and framing, of moment and relation with subjects-all of these, directly under the photographer's control, shape the end product. The way he controls it-what he decides to make it into-depends in the first instance on professional traditions and conditions of work. The kind of photograph he has learned to value and the possibilities for making them provided by the institutions he works in influence his decisions in general. Thus, for example, the short time periods magazine editors allotted to projects meant that photographers could not produce pictures that require lengthy acquaintance with the subject. Newspaper photographers do not, as a rule, make pictures that contain large blurred areas, because editors prefer pictures sharp enough to look good in newspaper reproduction (Rosenblum 1973).

A second influence on the image the photographer produces is his theory about what he is looking at, his understanding of what he is investigating. Saul Warkov says: "The camera is a wonderful mechanism. It will reproduce, exactly, what is going on inside of your head." That is, it will make the picture (given a modicum of technique) look just the way the photographer thinks it should look. Think of it this way: as you look through the viewfinder you wait until what you see "looks right," until the composition and the moment make sense, until you see something that corresponds to your conception of what's going on. Similarly, when prior to making the exposure you choose a lens and film, an f-stop and a shutter speed, you do so with the same considerations in mind. If you make exposures that look some other way than what makes sense to you, you probably will not choose them for printing or exhibition. Thus, what you expect to see and what, even if you did not expect it, you can understand and make sense of-your theory-shape the images you finally produce.
Since the skilled photographer can make the image look as he wants it to, and knows he can, photographers should be aware of the social content of their photographs and be able to talk about it at length. As a rule, they are not. One of the foremost recorders of the urban scene, Lee Friedlander, asked to verbalize the explicit social criticism his pictures seem to make, answered by saying, "I was taught that one picture was worth a thousand words, weren't you?" (Friends of Photography 1972:10). (And the recorder of the exchange adds that the audience of photographers and photography buffs burst into applause.) It is as though the criticism is there, but the photographer doesn't want to verbalize it directly, preferring to rely on intuition. In my limited experience with photographers, I have found that Friedlander's attitude, while not universal, is very common.
If the above remarks are accurate, then when social documentary photography is not analytically dense the reason may be that photographers use theories that are overly simple. They do not acquire a deep, differentiated and sophisticated knowledge of the people and activities they investigate. Conversely, when their work gives a satisfyingly complex understanding of a subject, it is because they have acquired a sufficiently elaborate theory to alert them to the visual manifestations of that complexity. In short, the way to change and improve photographic images lies less in technical considerations than in improving your comprehension of what you are photographing your theory. For photographic projects concerned with exploring society it means learning to understand society better. Insofar as sociology possesses some understanding of society (a very large if), then a knowledge of sociology, its theories, and the way they can be applied to specific situations might improve the work of both photographers and photographic sociologists.


A sociological theory, whether large scale abstract theory or a specific theory about some empirical phenomenon, is a set of ideas with which you can make sense of a situation while you photograph it. The theory tells you when an image contains information of value, when it communicates something worth communicating. It furnishes the criteria by which worthwhile data and statements can be separated from those that contain nothing of value, that do not increase our knowledge of society.

The work of social documentary photographers suffers then from its failure to use explicit theories, such as might be found in social science. This does not, of course, mean that their work embodies no theory at all. If they had no theory, they would have no basis on which to make the choices through which they produce their images. They have a theory, one which, because it is not explicit, is not available to them for conscious use, criticism, or development. Since they do not make explicit use of a theory designed to explore the phenomena they are interested in, they end up relying implicitly on some other kind of theory. The arguments that have attended the publication of some of the major works of obvious social import (e.g., Davidson's East 100th Street) indicate that the theories photographers rely on are, not surprisingly, lay theories, the commonplaces of everyday life in the intellectual and artistic circles they move in. Since photographers, for all their public inarticulateness, tend to be in touch (via their connections in journalism and art, and increasingly, through their location in academia), with contemporary cultural currents, they use the ideas and attitudes that are making the rounds in order to organize their own seeing.

That is probably overly harsh, since often enough photographers contribute images that help to shape those attitudes. Nevertheless, photographs of Harlem residents tend to revolve around such ideas as "Look how these people suffer" and "Look how noble these people are in the face of their suffering" (it might be argued that the latter was the twist Davidson relied on for the originality of his work). It is not that these things are incorrect or that for any reason they should not be said. But they are not sufficiently complex to sustain the weight of a real exploration of society, which will inevitably show that things are more complicated. In fact, the complications provide a great deal of the interest and points of active growth for social science thinking.

Training in social science, which presumably fills your head with social science theories, will not necessarily improve the social science content of your photographs. Knowledge does not automatically shape what you do, but works only when it is deliberately put to work, when it is consciously brought into play. Ruby (1972) argues that the pictures anthropologists take in the field are really vacation pictures, no different from the ones they take on any other vacation or that non-anthropologist vacationers take, focusing on what seems exotic and out of the way. Anthropological thinking does not affect the pictures. Photographic sophistication does. An unsophisticated photographer will produce a lot of isolated images while a sophisticated one will go after sequences of action.

Sociologists are probably like anthropologists. As they become more photographically sophisticated they will produce more interesting images, but not necessarily ones that have sociological content. Similarly, giving photographers a course in sociology or a list of suggested readings will not make their pictures sociologically more sophisticated. Learning some of what sociologists know will be necessary for improving the sociological content of their work, but it will not be sufficient.

How can sociological ideas and theory be brought to bear, in a practical way, on photographic explorations of society? The example of sociological fieldwork, as that has been described by a number of writers, (e.g., Lofland 1970; Schatzman and Strauss 1973), provides a useful model in the procedure of sequential analysis. I'm not referring to anything very esoteric, just to the procedure which allows you to make use of what you learn one day in your data-gathering the next day.


In some social science and photographic styles of work, you defer analysis until all the materials have been gathered. In a large-scale survey or experiment, the researcher can seldom change the way he gathers his data once he had begun; the inability to apply knowledge gained to the gaining of more knowledge is the price of standardized precision. (To be sure, one can apply the lessons of one survey or experiment to the next one, and workers in these styles usually do.) Photographers' failure to apply the lessons they learn at the beginning of a project to its later phases is more likely due to the photojournalistic emphasis on short intense trips to places one would not otherwise ordinarily be in, or getting the shooting done as rapidly as possible to cut down on expenses, and the great value placed on personal intuition, all of which have been elevated in some versions of photographic work to operating norms. (Like sociologists, photographers of course bring what they have learned in previous projects to bear on the next one.) Working in this style, photographers take advantage of their temporary presence in a situation to shoot a great deal, waiting until they have left the field to develop film, make contact sheets, and edit their results.


Fieldworkers work differently, in a way immediately adaptable to photographic projects. As they write up the descriptions and verbatim accounts that constitute their field notes, they simultaneously or shortly thereafter make preliminary analyses of that information (Lofland 1970; Schatzman and Strauss 1973). What is there in what they have recorded that they don't understand? How can they find out more about it? What ideas does it suggest about the organization they are studying and the people's experience in it? What patterns of interaction, of cause and effect, of interrelationship are suggested by what they now know? If the rest of what they observe is like this, what generalizations will they be able to make? Where should they look to find evidence that these preliminary ideas are wrong (or right)? In short, they develop tentative hypotheses about the object of their study, setting it in a context of theories and other data, and then orient their next day's observations and interviews along the lines suggested by the analysis. They try out different observable indicators of various sociological concepts. The concepts, embedded in theories, suggest links with other concepts and hence with other events observable in the situation, which can then be searched for, to provide both confirming and disconfirming evidence relevant to these provisional ideas. The analysis is continuous and contemporaneous with the data-gathering.

The photographer can do the same thing. To do so requires a longer time perspective than many photographic projects envision: certainly as much as the two years Davidson spent in Harlem, probably more than the seven months Winningham spent with wrestlers, or the couple of weeks that are even more common. To spend that much time requires establishing relationships with the people being photographed of a different order than those that photojournalists usually establish; it requires something akin to the research bargain sociologists make with the people they study. It means that the photographer has to find some way to support the long-term effort he is going to undertake.

Supposing that all this has been taken care of, let us consider how a sociologist photographer might go about such a sequentially organized project. He could begin by shooting almost anything he sees in the situation (the community, organization, or group), trying to cover whatever seems in a common-sense way to be worth looking at. The result is likely to be incoherent, visually as well as cognitively. The investigator will be learning how to work in the spatial arrangements and light situations in which what he is studying occurs. He will also be learning what is occurring, who the people are, what they are doing, why they are doing it. He learns the first by intensive study of his contact sheets and work prints; he should make plenty of work prints, in order to have something to study and hypothesize about. He learns the second in part in the same way. He looks at his work prints in a careful, detailed way, asking who all those people are and what they are up to. (Photographers tend to be satisfied with quick answers to these questions, and I think sociologists who would otherwise know better are just as likely to do that when they start working with a camera.) He should pay careful attention to details that don't make sense. For example, if people seem to be dressed in several distinctive ways, it pays to find out what status differences that marks, and then to ask in what other ways those groups differ. If people get into an argument which makes for a visually exciting image, it pays to find out why they are arguing. What is worth arguing about in that organization? What breach of expectations led to this argument? Do those circumstances occur frequently? If not, why not? BourkeWhite (1972:26), on photographing Ghandi, notes: "If you want to photograph a man spinning, give some thought to why he spins. Understanding is as important for a photographer as the equipment he uses. In the case of Ghandi, the spinning wheel is laden with meaning. For millions of Indians, it was the symbol of their fight for independence."

The photographer pursues these questions with his camera, but also by asking people about what he has seen and by observing closely and listening carefully as the everyday activities of the group go on around him. He should not keep away from the people he is working with, shooting from a distance with a long lens, but rather should get up close and establish a working relationship with them, such that they expect him to be there and accept that he has some sort of right to be there which he will probably exercise most of the time. (Aside from the visual considerations, photographers doing this kind of research might want to use a wide-angle lens, perhaps 35mm, as standard equipment, because it will force them up close where they ought to be.)

The photographer can also get more data by showing people the pictures he has already taken. He probably will have no choice, because people will want to see what he's up to. This will give him the chance to use the photo elicitation technique Collier (1967) describes so well: showing the pictures to people who know the situations under study and letting them talk about them, answer questions, suggest other things that need to be photographed, and so on. 5

If the photographer has some sociological ideas available, he can apply them to these more or less commonsense questions and answers. Much of what I've described so far is only what any reasonable curious person might want to know. Nevertheless, basic sociological theory is involved, one compatible with most varieties of sociology in current use. Let me put it in the form of a list of questions to be answered in the field, cautioning that the answers don't come all at once, but through a process of progressive refinement and constant testing against new information. This formulation of the questions a sociological-photographic study could usefully orient itself to is not original; it has been heavily influenced by Everett Hughes (1971).

(1) What are the different kinds of people in the situation? They may or may not look different; they will certainly be called by different names.
(2) What expectations does each kind of person-members of each status group-have about how members of other groups ought to behave? What are the recurring situations around which such expectations grow up?
(3) What are the typical breaches of those expectations? What kinds of gripes and complaints do people have? (A complaint is a sign of a violated expectation; "He's supposed to do X and he hasn't.")
(4) What happens when expectations are violated? What can people do to those who do the violating? Is there a standard way of settling these conflicts?

These questions put in a commonsense way ideas integral to almost any sociological analysis. (1) refers to what a sociologist might call status groups; (2) to norms, rules, or common understandings; (3) to deviance or rule violations; (4) to sanctions and conflict resolution. The advantage of the translation is that these concepts are linked in such a way that if you identify something you have seen as an instance of one of them you then know that you ought to look for other things that will embody the ideas it is connected to in the theory. If, for instance, you see someone reward or punish someone else, the theory directs you to look for the expectations that have been violated in this case, and for the status groups to whom those expectations apply. Anyone exploring society photographically can ask these questions, both visually and verbally. Each day's data provide some provisional answers and some new questions, both discovered by careful inspection and analysis of the material.

The photographic investigator can supplement his visual material with a running verbal record. Depending on his intentions, this might be a full set of field notes such as a sociologist doing a conventional field study would keep, complete with verbatim conversations, or a record of a few outstanding thoughts and remarks. Some photographers (e.g., Winningham and Lyon) have tape recorded interviews with the people they photograph. Some (e.g., Owens) have recorded the responses of people to their photographs.

As the work progresses the photographer will be alert for visual embodiments of his ideas, for images that contain and communicate the understanding he is developing. That doesn't mean that he will let his theories dominate his vision, especially at the moment of shooting, but rather that his theories will inform his vision and influence what he finds interesting and worth making pictures of. His theories will help him to photograph what he might otherwise have ignored. Simultaneously he will let what he finds in his photographs direct his theory-building, the pictures and ideas becoming closer and closer approximations of one another. Like the sociological fieldworker, who finds much of his later understanding latent in his early data (Geer 1964), he will probably find that his early contact sheets, as he looks back through them, contain the basic ideas that now need to be stated more precisely.

The photographer, like the sociologist who builds more and more comprehensive models of what he is studying (Diesing 1971), will arrange the visual material into the patterns and sequences that are the visual analogue of propositions and causal statements. He will consider the problems of convincing other people that his understanding is not idiosyncratic but rather represents a believable likeness of that aspect of the world he has chosen to explore, a reasonable answer to the questions he has asked about it


SOME COMMON PROBLEMS

Whether they start as sociologists or photographers, anyone who undertakes the kind of project I have just described will run into certain problems, which are common both in being frequent and ubiquitous and in being shared by the two vocations. In some cases, sociologists have ways of dealing with problems that photographers might find useful; in others, the way photographers deal with those problems will throw a new light on sociologists' troubles.

Truth and Proof

Insofar as a photograph or group of them purports to be "true," the particular meaning of that ambiguous claim needs to be specified. Once we know the kind of truth a picture claims, we can assess how far we accept the claim and how much of the statement it makes we want to believe.
Photographs (barring those that have been obviously manipulated to produce multiple images and the like) minimally claim to be true in that what they show actually existed in front of the camera for at least the time necessary to make the exposure. Photographs in the social documentary style claim more than that, presenting themselves as pictures of something that was not done just for the photographer's benefit, but rather as something that occurs routinely as part of the ordinary course of events. Or the photograph suggests that what we see is, if not ordinary, characteristic in some deeper sense, portraying some essential feature of the phenomenon photographed. When people speak of a photograph having "captured" something, they generally mean that it displays some such characteristic feature. Frequently, though not always, the photograph suggests that what it shows, while characteristic, is ordinarily hidden from view, so that we might never know its particular truth if the photographer did not show it to us. Many photographers make no such claims, at least explicitly, preferring to avoid the responsibilities that accompany the claims by describing their pictures as containing only the truth of "how it felt to me." This makes the photograph the visual analogue of something like a lyric poem, its author's sole responsibility to have rendered honestly his own feelings and responses. Such work can be interesting and moving; we often feel that, because we trust and feel some empathy with the lyricist's sensibility, we have learned something about the world from his response to it. The lyric poem or photograph need not give us that bonus, however, and its maker needn't satisfy any requirements of truth or objectivity.


Photographers frequently find themselves troubled because, after they have shown us some way of seeing a part of society, someone else accuses them of not having told the truth. Perhaps the photographs are not what they claim to be: though they appear to be "candid" portrayals of everyday events, the people or objects in the picture never really appeared that way, and only did so at the time of the photograph because the photographer posed them (as in the case of the flag-raising at Iwo Jima or the controversy over Arthur Rothstein's picture of a skull on parched Dust Bowl earth (Hurley 1972:86-92), where opponents said he had made an old skull appear to be the product of the recent drought). Photographers often feel the accusation that they set up a shot, rather than photographing something that occurred naturally, to be damaging. When they do, they reveal the degree to which they are claiming something more than subjective truth for their work.
In a commonsense way, people make judgments about that threat to the validity of a photograph (to paraphrase Donald Campbell's useful notion of the threat to the validity of a hypothesis). We may base the judgment on evidence contained in the photograph, recognizing that we have seen similar things elsewhere, so that their existence is not in question; the photographer has simply called our attention to something we already know. The photograph may have been made in a place so public and accessible to independent checks that we reason the photographer would not fake something whose phoniness could so easily be discovered. We may rely on the established reputation of the journal the photograph appear in, being sure that Life would not risk its reputation for accuracy just for the sake of this one picture. How we establish the credibility of a photograph is a problem in commonsense reasoning I won't pursue further here.


When the validity of the individual photographs is not in doubt, a more serious question about the "truth" of a presentation remains. Couldn't someone else have photographed the same people, places or events and produced a quite different statement about that social reality? Any collection of photographs is a selection from a much larger population of photographs that have been or could be taken, and the answer to the question is necessarily yes, that reality could have been presented in another way. I don't know why photographers are as sensitive as they are about this, since they have a simple counter available to the accusation of "bias." The answer lies in distinguishing between the statement that X is true about something and the statement that X is all that is true about something. Thus, Neal Slavin's photographs of Portugal prompted one critic to complain that he couldn't believe that, as this portfolio suggested, no one ever smiled in Portugal. If photographs indicate that other phenomena, even though not central to the statement being made, exist, much of this difficulty could be avoided. Sociologists typically plaster their work with such caveats. Statements so qualified lose something in dramatic impact, but they gain in credibility over the long run; you can choose which you'd like, but you can seldom have both.

Sampling

Another version of the same problem arises when, having assured ourselves that the photographs are valid and that, while they claim to be true, they do not claim to be the whole truth, we ask: if we had gathered our data at some other time, or from some other part of the universe our assertion applies to, would we get essentially the same result? Put it another way: if I know what I do about these people and places at this time, what else can I be reasonably sure I know about? Sampling problems have two aspects: (1) what procedures shall I follow to maximize the generality of my findings? and (2) how can I convince others that my findings have that generality? The first question is procedural, the second rhetorical. Social scientists often deal with the two questions simultaneously. They use a certified technique whose logic is well known; by asserting that the appropriate procedure has been used, they assure readers that their conclusions follow logically. For photographers, the two questions more frequently arise separately.
Social scientists deal with threats to the generality of their propositions by a variety of sampling techniques. If they are concerned with whether certain quantitative distributions or relationships found among those they have observed approximate those in the larger universe from which their observations were drawn, they may use some version of probability sampling. If they want to make sure they have covered all the major aspects of a group's activities or of a social organization, they may rely on what Glaser and Strauss (1967) have called theoretical sampling, choosing units for observation because some theory suggests they would be strategic.

Photographers are seldom concerned with quantitative generalizations, or with covering some theoretical map adequately. But they often present their material in a way that suggests they believe that what they show us applies to a far wider area and population than the one they have covered, that were we to look at a different part of the same whole, we would see more of the same. I don't know what procedures photographers use to assure themselves about these matters. Sociological fieldworkers use some simple procedures that would serve the double function of maximizing generality and thus responding to such queries, and simultaneously enlarging the possibility of getting unanticipated and possibly exciting material (both sociologically and visually). Following some of these suggestions might produce a lot of dull pictures, but so do most procedures; exciting and informative photographs are always hard to come by.

Fieldworkers may use crude time-sampling devices: checking up on someone or someplace every half-hour, or on different days of the week, or different times of the year. Some avoid "leaving things out" by attaching themselves to one person at a time and following that person through his entire daily (and nightly) round. They may ask people under study who else they ought to talk to or observe. As they become aware of categories or situations that deserve special study, they can systematically choose some to observe or they can observe all of them. Fieldworkers follow the discipline of recording everything they see and hear while making these observations.

Photographers could do all of these things, but they would need to observe some discipline equivalent to incorporating everything into the field notes, for a photographer's data do not exist unless they expose some film. In following someone around for a day, they might for instance adopt some such convention as exposing at least one roll of film every hour or so, adapting the time period to the character of what they were observing. They would thus avoid waiting until "something interesting" happened, and increase the chance that things that don't as yet fit into the photographer's developing understanding would nevertheless get into the record. They might similarly photograph certain activities or places on some schedule that interferes with their tendency not to shoot what does not seem visually interesting. Any kind of theory of the kind discussed earlier would likewise direct the photographer to things his intuition and visual sense might not call to his attention. Remember that theory is itself a sampling device, specifying what must be incorporated into a full description.


Shooting what seems interesting usually satisfies the photographer's need for a method. However, they often realize, if they are sensitive to their own work, that they are producing essentially the same pictures in a variety of settings, because their notion of what is visually interesting has become divorced from the social reality they are working in. If they are not sensitive to that possibility, others might point it out. A technique that breaks up their established visual habits guards against this. In addition, photographers often find that they are slow to discover and shoot things they later realize they need for a more complete visual understanding. The same techniques of randomized and theoretically informed sampling may help. The object of all this is not to turn photographers into sociologists or enslave them in mad sociological rituals, but rather to suggest how sociological tricks might solve problems of photographic exploration.


Sociologists try to convince their readers that generalizations from findings are legitimate by indicating that they have used a conventionally approved technique. The scientific community has already inspected the logic of that technique, so it is sufficient to indicate that it has been appropriately used. Readers who accept that convention are automatically convinced.
No photographer uses such standardized devices, and I'm sure that none would be interested in pursuing such techniques as probability sampling. They have their own devices, however, worth exploring because these produce conviction in the viewers of photographic work similar to that produced by sampling designs in sociological readers. Since sociological procedures are, to quote Campbell again, "radically underjustified," it is worth considering photographers' methods, even though they may appear even more underjustified to sociological readers.

The chief device photographers use is to identify their photographs by place and sometimes by date. The photographs in Frank's The Americans are identified simply by a generic organizational type and a town: "Bar-Gallup, New Mexico," "Elevator-Miami Beach," "Bank-Houston, Texas." Dennis Stock's (1970) California Trip identifies the individual images by town and/or neighborhood: "Sunset Strip," or "North Beach, San Francisco." These labels, coupled with a reiteration of themes, so that one sees the same kind of place or thing or person from half a dozen widely scattered places in the country, imply the conclusion that if you can find it in that many places, it is really very widespread. Thus, when Frank shows you luncheonettes, diners, and coffee shops from Indianapolis, Detroit, San Francisco, Hollywood, Butte, and Columbia, South Carolina, all of which share a gritty plastic impersonality, you arc prepared to accept that image as something that must be incorporated into your view of American culture. The logic of this deserves further analysis, since it is convincing (there are other such devices which need to be described and analyzed).

Reactivity

The problem of the reactivity of data-gathering procedures is very similar in ethnographic and photographic work. Does the sample of behavior observed and recorded accurately reflect how people ordinarily act or is it largely a response to the observer's presence and activities? Both sociologists and photographers frequently deal with this by cultivating the art of being unobtrusive. Many people know how to manipulate their bodies and expressions so that, in the absence of any reason to pay special attention to them, the people they are observing ignore them; how they actually do this is not explicitly known, and deserves investigation. It is probably easier to be unobtrusive in public places where you are not known as an investigator and it may or may not be easier if you are carrying a camera. In many situations carrying a camera validates your right to be there; as a tourist, as a member of the group recording the scene for their purposes, or as a representative of the media. Under many circumstances, observing or photographing is commonplace and expected; many other people are doing it. Your presence does not change anyone's behavior since observers and photographers are part of the situation. You should, of course, include their presence in your observations and photographs.

In many situations, the people being observed are engaged in activities of considerable importance to them and cannot change what they are doing for an observer's benefit even if they would like to. Reactivity depends on the freedom of those observed to respond to the observer's (or photographer's) presence. If they are enmeshed in the constraints of the social structure in which they carry on their normal activities, they will have to carry on as they ordinarily do for whatever reasons cause them to do that ordinarily (Becker 1970). They may be well aware that they are being observed or photographed, but not be free to change what they do. Photographers routinely make use of this possibility. I once watched Michael Alexander photograph a woman fighting with her small child in a playground. Alexander was practically on top of her, but the child was kicking and screaming and, though she had no idea who he was, she felt she had no choice but to deal with her child despite the unwelcome recording going on.
A third solution recognizes that the reactivity often reflects fears about what will be done with the information or photographs. If the observer gives evidence that these will not be used to harm the people he is observing, they may decide to ignore him, or to cooperate, for instance, by pointing out things that need to be investigated or photographed, or by keeping him up to date on things that have happened while he was not around.


Photographers make use of a fourth possibility that sociologists seldom employ, though it is the chief element in studies of experimenter bias and similar problems. They encourage reactivity and make it the basis of their exploration of people and events. The photographs become a record of their relationship with the people they photograph, and the reaction of the people to being photographed becomes the chief evidence used in analyzing them. Sociologists make use of this possibility when they look at the difficulties of gaining access as revelatory of the social structure to which access is sought (e.g., Gardner and Whyte 1946).


Getting Access

Sociologists have increasingly worried about the conditions under which they will be allowed to gather data and then make their research results public. Science requires that data and operations be open to public inspection and independent verification. Unconstrained, scientists would (and should) make all their data public. But they are constrained by both legal and moral considerations from doing so, and ordinarily take substantial precautions to avoid harming anyone by revealing who furnished information for or are the subjects of research. They may simply change the names of people, organizations, and places, or use elaborate coding procedures to preserve the anonymity of survey respondents.

People sociologists write about seldom sue them (though my colleagues and I were once threatened with a libel action by the administrator of an organization we studied). Consequently, they worry more about ethical than legal problems. Though a substantial literature debating these problems has grown up, the situation is confused and sociologists do not agree on procedures or relevant ethical principles. They tend to agree on generalities-"We should not do harm to the subjects of our research"-but not on the application of such crucial terms as "harm." To take one example: Are organizations, and especially such public ones as governmental agencies or schools, entitled to the same privacy as individuals, or is not social science research part of the public review to which they are necessarily subject? Another: Where do you draw the line between inconvenience or embarrassment and substantial harm?

Photographers have been considerably more interested in legal problems. When they make simplified analyses of the problems they explore, they can take an equally simplified view of the ethical problems. Having no trouble telling the good guys from the bad guys, they have not had to worry so much about ethical questions. If their work hurts the bad guys on behalf of the good guys-well, that was the point. But they have had to worry about being sued for invasion of privacy, and libel. The law here seems to be as ambiguous as the ethical standards of sociologists. Photographers know they can be sued and often take the ritual precaution of having people sign standard release forms, though these may not be as useful as supposed. 6 They also try to maintain friendly relations with the people they photograph, in much the same spirit as the advice I heard given to medical students: if you are good friends with your patients they won't sue you for malpractice. Alternatively, they rely on this being a large, differentiated society in which it is relatively unlikely that anyone will see the picture of him you put in a book or exhibit.
Everett Hughes' (1971) idea of the research bargain provides the terms for a useful comparison. What bargain do investigator and investigated make? In both photographic and sociological investigations, it is fair to say, the people investigated probably do not know what they are getting into. They may give their consent, but it is not an informed consent. From an ethical and perhaps a legal point of view, the bargain is not fully valid. Sociologists are generally very cautious about this, at least in public discussion, and I think they might consider seriously a view more common among photographers: people can and should take care of their own interests and once the investigator has honestly described his intentions he has fulfilled his obligations. I don't propose that we accept this view uncritically, but we might think hard about why we should not. Journalists have long operated with a different ethic and there is perhaps as much reason to adopt their practice as that of physicians, which has tended to be the one sociologists orient themselves to.


Photographers have probably taken a tougher line because they can't use some of the devices sociologists do. Unless you block out faces and other identifying marks, everyone in a photograph is identifiable and there is no possibility of preserving anonymity. That is the strength of the medium, and no one would sacrifice it for ethical considerations. The strength of photographic work may not depend on the people and organizations studied being identified specifically, since the implicit argument is that what you see is characteristic of a large class; so the people in the individual prints are in effect anonymous, though they might be known to some who see the pictures and others could conceivably find out who they are if it seemed important. (But see Alwyn Scott Turner's Photographs of the Detroit People, in which a great many people photographed are not only named but their approximate addresses are given, too.)

The other aspect of the photographer's situation that leads him not to worry so much about ethical considerations is that when he is not photographing anonymous people who will be made to stand for some more general aspect of the human condition he is usually photographing people who, because they are public figures, expect to be photographed and only complain when it is grotesquely overdone, as in the case of Jacqueline Onassis. These people epitomize the rationale I mentioned earlier: perfectly capable of defending their own interests, they accept their photographic burden as one of the costs of being a public figure, whether they like it or not.

Both these strategies offer possibilities for social researchers. Sociologists frequently disguise names of people and organizations without thinking why, and might often be able to identify them, particularly when what they have said or done is no more than ordinarily discreditable and when (as is inevitable in social research) a long time elapses between getting the information and putting it into print. Studs Terkel has done that in his books on Chicago and on the Depression to good effect and without doing anyone harm.

Similarly, we might treat public figures as just that, justifying our observations, interviews, and quotations on the grounds that we are entitled to them as citizens and need no special social science warrant for our actions. A good example appears in a study by a combined legal and social science research staff of public access to information (Northwestern University Law Review 1973). As part of an elaborate experiment, researchers visited a number of public offices in search of information to which their access was guaranteed by law. Information holders often refused them or evaded their requests with transparent devices; the researchers in providing evidence for their conclusions, described their encounters with public officials, identified by name and office. I see no reason why that device should not be used more often than it is.

Concepts and Indicators, or Ideas and Images

Sociologists tend to deal in large, abstract ideas and move from them (if they do) to specific observable phenomena that can be seen as embodiments, indicators, or indices of those ideas. Photographers, conversely, work with specific images and move from them (if they do) to somewhat larger ideas. Both movements involve the same operation of connecting an idea with something observable, but where you start makes a difference. Granting, and even insisting as I already have, on the conceptual element in photographs, it still is quite different to start with something immediately observed and try to bend ideas to fit it than to start with an idea and try to find or create something observable that embodies it. Sociologists have something to learn from photography's inextricable connection with specific imagery.

Many sociological concepts, whose meaning seems intuitively clear, would he very hard to portray visually. Consider the notion of status integration. Defined as a congruence (or lack of it) between two or more indicators of social rank (education and income, for instance), its human meaning seems obvious. A man who made $100,000 a year but had never finished grade school would, we can imagine, have troubles another man with the same income who had completed college would never know. Does it have a visual counterpart? Can we imagine what a person in either of those states would look like, what we might see him doing, what his possessions and environment would consist of? The answer, to both questions, is probably no.

We cannot imagine the visual counterpart of status integration, I think, because the concept has been defined by the rules for calculating a status integration score from numerical indicators of specific ranks. The human meaning of the concept has been left to be evoked intuitively from the label applied to the results of that operation. As a result, no one can be sure what an instance of status integration would look like and thus no one can photograph it.

Obviously, every sociological idea need not be connectable to a visual image to be valid or useful. On the other hand, consider this. Some sociologists describe a basic problem of empirical research as one of finding empirical indicators (things observable in real life) to measure a concept whose meaning they have already decided. A sizable literature discusses the logic by which the two can be defensibly connected. But, as the example of status integration suggests, a third element is involved: the basic imagery we intuitively supply to fill out the meaning of an abstract concept operationally defined. We seldom consider the logic by which we connect concepts and indicators to that basic imagery, or the procedures by which we can develop that imagery explicitly and connect it defensibly to concepts and indicators. While, to repeat, sociological ideas needn't evoke a clear visual image to be defensible, considering the processes by which photographic imagery arises may help us understand what is involved.

The gap that develops between concept and indicator, on the one hand, and basic underlying imagery, on the other, is nicely illustrated by a device Blanche Geer uses in teaching fieldwork to graduate students. They are given to talking in rather grand theoretical terms when asked to describe what they have seen, and she counters this by asking if any of them have observed a status (or norm or social structure or whatever). When someone claims to have observed such a thing, she asks what it looked like, what it said, how it acted. She thus hopes to make students understand that such terms are shorthand for a class of observable phenomena that can be described, and have no more reality or meaning than they get from the collection of phenomena so described and the resemblance's among them.

The imagery underlying a sociological concept implies, if it does not state explicitly, a picture of people acting together. It may picture them engaged in familiar forms of social interaction, or it may imply a more mechanistic vision (as when people are conceptualized as members of an aggregate rather than an interactive group, in which case the imagery may be of something like social molecules engaged in an analogue of Brownian movement). In either case, the concept and its indicators evoke (even when they use the language of operationally defined variables) an image of social life. The fidelity of that imagery to the realities of social life is, as Blumer (1969) has emphasized, an important issue in assessing the utility of a concept.

When the imagery underlying a concept is explicit, it can more easily be criticized and revised. Durkheim (1951) for example, gives very explicit and vivid descriptions of the collective and individual States which he defines abstractly as embodying the theoretically defined quality of anomie. We can easily judge for ourselves how well the abstract concept and the empirical indicators mesh with the imagery. Where the underlying imagery is left implicit, the reader invents his own and the critical assessment of that relationship tends not to occur.

We might expect, as a result, more dispute over the meaning of theoretical concepts than there is, because differing underlying images lead to a different understanding of a concept's meaning, use, and appropriate measure. One reason for the lack of dispute is the sociologist's tendency to discuss concepts in a purely verbal and logical way divorced from any close relation to empirical materials. When they do that they play on the underlying imagery without taking responsibility for it. Several generations of psychologists have played that game with the concept of intelligence, defining it operationally, saying "Well, let's call it X" when its validity was questioned, but never calling it X because they would then lose the meaning imparted by the imagery associated with "intelligence." (They thus paved the way for the excesses of Jensen, Herrnstein, and Shockley.) If we cannot imagine or discover a visual image that embodies our understanding of a concept, we might take that as a warning that the concept is not explicitly related to its underlying imagery. Looking for an appropriate visual image might help clarify the relationship.

Photographers, of course, do not have this problem. They work in the opposite direction, needing to find concepts that adequately convey what is important in what they give us to see, the explicit conceptualization working for both photographers and viewers to provide a framework for their joint work of making sense of what they see. I've already discussed how the failure to use explicit concepts and theories hampers the development of photographic analyses and how sociological ideas might be brought to bear on the development of photographic projects. What photographers do very well, however, is to refine over a period of time the image they create of something. They may photograph people, places, and situations again and again, seeking to make the resulting image express more clearly, concisely, and unambiguously their basic understanding of those things. They tend to approach this task visually, stripping away extraneous elements so that the statement the image makes communicates its substance efficiently and emphatically to the viewer.

Sociologists might well work at the job of continuously refining not only their concepts and measures but also their basic imagery, relying on that refinement more than they have to clear up theoretical and technical muddles. Blumer has often recommended something like this, though he hasn't been very explicit about what is involved, so his advice sounds mystical. I don't at this time have any less mystical and more specific suggestions. The basic idea, however, is to clarify how you think things really are, using the imagery you develop as a touchstone against which to test concepts and indicators as these develop.

Boundaries, Limits, and Framing
No intellectual or artistic enterprise can include everything. Scientific studies tend to make clear that they have a limited area of responsibility, that they are only studying these phenomena, this area, the relationship between these variables and those; while other things may be important too, they will be left out, since you can't study everything at once. Scientists often contrast their practice in this respect with that of artists or novelists who they caricature as striving to include "everything" in their works, as though most artists were super-realists or as though even super-realists actually included "everything" or thought they did. In fact, artists leave things out too. But their selectivity is more conscious, and they often use as an artistic resource the necessity to choose between what will be included and excluded. They make the selection itself an artistic act. They rely on the viewer's tendency to supply in imagination what is not present to make allusion work in the total statement, so that what lies beyond the frame becomes an integral part of the work. For photographers, "framing"-choosing what will go inside the bright line of the viewfinder-is one of the key decisions.
The choice of the boundaries of a study has an enormous effect on the results. For social science, it has among other things a strong political effect. What we choose not to study becomes a given in our research. We rule out the possibility of taking its variations into account (though we can of course focus on them in some other study, so the tendency I am talking about is only a tendency, not a rigid constraint). We may thus come to regard what we choose to see as fixed as being in fact unchanging. We see this tendency at work, for instance, in any statement which suggests that an organization must do some particular thing (e.g., satisfy some particular need or requirement) if it is to survive. The statement is misleading unless we interpret it as shorthand for the cumbersome proposition that it will change from its present form of organization and level of performance in various ways if the particular need or requirement is met at some other level or in some other way than that specified. When we put it that way, we recognize that survival, which the simpler statement treats as a given, can be made problematic and variable.

The political effect comes about when we take what we have defined, for scientific convenience, as unchanging, as in fact, unchangeable. We thus, implicitly or explicitly, suggest to those who think that some particular change is the way to solve a pressing problem, that their solution is utopian and unworkable. What we are really saying, in such a case, is that the phenomenon in question can only be affected by changing something so difficult to change that only extraordinary effort and power can accomplish the feat. The mobilization of effort and power might be accomplished, if only in a way that the analyst might think unlikely or distasteful (e.g., violent revolution).


Likewise, when sociologists fail to consider some people and some aspects of a situation and do not gather data about them, they forego the possibility of finding out that some things said by or about those people are not true, that their informants' descriptions of their own actions may be self-servingly misleading. For social scientists, this choice usually results in studying subordinate echelons in an organization or community, while taking the descriptions by superiors of their own activities as adequate and trustworthy and therefore not needing any investigation. This lack of scientific skepticism is a political choice and has political consequences (Becker 1967; Blumer 1967; Becker and Horowitz 1972).

Since photographers seldom produce explicit analyses of social problems, they are less likely to confront this problem directly. But their idea of who should be photographed and who should not may have the same consequence's as the sociologist's decision about who is to be studied, the photographer thereby giving us great informational detail about some people, and suggesting that others either do not exist or can be filled in from the viewer's imagination. How, for instance, would Hine's documentation of the problem of child labor have been affected had he included among his portraits of exploited children portraits of the men and women who owned the factories, profited from that exploited labor, and lived in extravagant luxury on the profits? It might have given a more damning indictment of the entire system, though it is questionable that his work would then have had greater effect. One could also argue that the machines and factory buildings present in his pictures convincingly evoke the owners and their power (though not the luxury of their lives), or that other photographers provided that material, e.g., Steichen's (1963:31) portrait of J. P. Morgan.

Another aspect of framing is that we can either include all of what we do show within the picture's frame, and thus indicate that it is self-contained, or include parts of things that extend beyond the frame and thus evoke the world into which they extend, or things that stand for and evoke worlds and situations which lie beyond. Portraits, for instance, can contain all of the person's body and thus indicate that it is not necessary to know more, or they can contain only parts and thus indicate that there are other parts the viewer must supply from his imagination. Likewise, a portrait can contain some chunk of the person's ordinary environment-an artist's studio, a scientist's laboratory which evokes a world of activity not pictured, but there. Or it may simply show some setting (home or whatever) in such a way as to suggest more about the person. André Kertész (1972:118-119), for instance, has a portrait of Mondrian that faces a picture of Mondrian's house, which arguably conveys a more Mondrianish spirit than the portrait of the artist himself.

In any event, photographers do understand and use what lies beyond the portion of reality they actually show. In this they differ from social scientists who prefer not to discuss explicitly what they cannot claim to have studied scientifically. In that sense, social scientists make themselves ignorant about matters that lie beyond their frame, ignoring even what they do know by casual observation or in some other informal way. Instead of building such partial knowledge into their analyses, they rely on time-honored verbal formulae (e.g., "all other things being equal") to limit and frame their analyses. These formulae, like legal formulae, have been revised and refined so as to say exactly what is meant, what is defensible, and no more. A large number of these conventions exist, part of the rhetoric of contemporary science.

In any event, when social Scientists fail to deal with the reality that lies beyond the frame they placed around their study, they do not get rid of it. The reader, as with photographs, fills in what is hinted at but not described with his own knowledge and stereotypes, attaching these to whatever cues he can find in the information given. Since readers will do this, whatever verbal formulae are used to attempt to evade the consequences, sociologists might as well understand the process and control it, rather than being its victims .

Personal Expression and Style

Sociologists like to think of science as impersonal. However, they recognize that people work differently, that some have easily recognizable styles of work, that some work has an elegance missing in other research. In short, they recognize a personally expressive component in sociological research and writing. They seldom discuss that component (I suppose because it contradicts the imagery of impersonal science). When they do discuss it, they usually describe it as a flaw. For instance, critics frequently complain of Erving Goffman's jaundiced view of the world, of modern society, and especially of personal relationships. They characterize that view as overly calculating, as cynical and even as paranoid. Similarly, some critics of so-called "labelling theory" criticize it for being overly skeptical about established organizations, their operations and records.

Both Goffman and labeling theorists have the elements these criticisms single out. So does every other theory and style of work. The critical analysis errs only in suggesting that some theories and studies have such components while others are properly impersonal, as befits scientific activity. But Blau and Duncan's (1967) study of the occupational structure of the United States, to take a random example, likewise contains a personally expressive element, both in its view of the nature of people and society and in the way it handles and presents data, even if we see that element minimally, as a nonsharing of the Goffman view. The style of scientific impersonality is also a style.


Photographers typically accept responsibility for the personally expressive component in their work as a natural accompaniment of its status as art. Accepting that status also allows them the quasi-mystical retreat from analyzing the social components of their work and the emphasis on intuitive inarticulateness I criticized earlier. Nevertheless, they understand something sociologists need to learn more about, so they can work with it consciously and control it.

We can approach the serious analysis of the personal component in sociological work by looking at specific devices through which it is expressed. There is a dictionary of the expressive language of photography yet to be compiled; at present, I can only find occasional ad hoc discussions.

Here is an example of the stylistic devices that express the personal component in photographs. Paul Strand (1971) is famous for his portraits of peasants from all over the world: Mexico, Morocco, Egypt, Romania, the Gaspe, the Hebrides. The portraits overwhelmingly convey an attitude of respect for the people portrayed, describing them as strong, sturdy, enduring, good people who have the traditional virtues despite the difficult circumstances of their lives. This is quite a different description from that of ethnographers as various as Tax and Banfield, who depict people who are meaner, more cunning, more spiteful. Strand has chosen to portray them that way. He has not simply conveyed the reality of peasant life. He conveys his view by habitually photographing his subjects at eye level, directly facing the camera, thus treating them as equals. He does not suggest that he has caught them in an unguarded moment; on the contrary, he has allowed them to compose themselves for the occasion, to put their best foot forward. The stability implied in their formal postures, the honesty suggested by the openness with which they gaze into the camera, all help to suggest peasant virtues. Likewise, by photographing them in natural light and utilizing a wide tonal range, Strand conveys an attitude that respects their reality, that makes them look fully human.

Frank Cancian's (1971) photographs of Mexican peasants use different devices to convey a view of peasants which is (not surprisingly, since Cancian is himself an anthropologist) much more like that of earlier ethnographic descriptions. His Zinacantecos occasionally show the nobility Strand emphasizes, but more frequently seem less noble and more human. They grin, smile slyly, bargain shrewdly, drink hard. The photographs view them from a variety of angles, show them in blurred motion, in a variety of light conditions, all of which express somewhat less respectful distance and some-what more knowledgeable familiarity than Strand's pictures. The difference in knowledge of and attitude toward the people being photographed is conveyed by the choice of topics too, of course, but the stylistic elements play an important role.

I'm not sure where we might find the expressive devices characteristic of sociological work. One place is in the use of adjectives. Sociologists frequently, perhaps in an attempt to achieve a little literary grace, apply adjectives to the people and organizations they write about, these adjectives implying judgments and generalizations not justified by the data they present or required for the scientific points they are making. A variety of other devices known to literary analysts likewise convey attitudes and moral evaluations. Goffman, for instance, often achieves ironic effects by using perspective by incongruity, and many people use a Veblenesque deadpan translation of evaluative statements into mock-objective academese to the same end.


Sociologists use a variety of devices, interestingly, to hide the personal attitudes, evaluations, and other components in their work. Chief among these are the incessant use of the passive voice and the first person plural to blur recognition of what is obvious: that one person is in fact responsible for the research and results being reported. Even more interesting to me is how do various styles of handling quantitative data contribute to a rhetorical effect of impersonal fact? What are the aesthetics of tabular presentation? These questions, to which I have no answers, lay out an area of work still to be done.

CONCLUSIONS

This paper is made up of notes from work in progress, and what I have said is necessarily preliminary and incomplete. The kind of work it intends to encourage barely exists as yet, though the common and converging interests of social scientists and photographers, often in the same person, suggest that we don't have long to wait. I hope the paper will provoke further discussion and work on the problems it proposes.